All posts by Alex Kvenvolden

All three Stride BRT lines are set to launch in 2024. While there are some refinements on the north side of I-405 with expansion of the express toll lanes and new inline stops, BRT benefits will be more significant on the south side of I-405, which is getting express toll lanes for the first time (compared to the frustratingly ineffective HOV2 lanes on the corridor today).

While freeway improvements would go a long way to improving express bus service in this corridor, there is still a problem of how service is organized and routed currently. Today’s bus service is a mix of remnants of the old service pattern from the early 2000s (when Bellevue buses went to Federal Way and South Hill) and new route 567 service which takes advantage of timed Sounder transfers and the new HOV ramp between I-405 and SR 167. (In the spirit of full disclosure, 3 out of 8 of these points affect me very directly, and were key motivators in my writing this post)

Here are some issues with the system as it stands now:

• Auburn to Bellevue service is part of the 566 rather than the 567, so riders are faced with the uncomfortable decision to take the 566 or Sounder and transfer to the 567, or take the 566 all the way, unsure of what will get them there faster.
• Current 567 trips (which connects 1 to 1 with Sounder trains) very often fill up in Kent after the train arrives, requiring passengers to choose between waiting for the next 567 (20 minutes later) or dealing with the slower 566.
• Service in Renton is oriented around picking up many people along local stops and serving the Boeing plant, without requiring passengers to transfer, no matter where or from they are going. This is a problem because it slows service substantially, despite relatively few people utilizing these stops compared to major stops like Renton TC.
• Specifically, Renton-Bellevue service is interlined with “pseudo 10-minute” headways, where the 560 and 566 are coordinated to run every 10 minutes from Renton to Bellevue. The problem with this is that the 560 serves both freeway stations on I-405 (whereas the 566 serves none, usually), so 560 takes longer. This requires schedule tricks when going northbound (560 leaves a few minutes earlier than 10 minutes after the 566 to get to Bellevue at the same time). And southbound in the evening, this trick doesn’t even work at all (560 takes more than 10 minutes longer than the 566, so the next 566 trips are scheduled to leave Renton TC before the previous 560 does).
• 2/3 of northbound morning Renton-Bellevue service comes directly from Kent on SR-167 (route 566), and as this gets congested with rush-hour traffic, it often severely delays trips from Renton before they even start (a scenario that most major rush-hour service avoids, such as Federal Way to Seattle on the 577, Renton to Seattle on the 101, and Kent to Bellevue and Seattle on the 567 and Sounder). This is further worsened when getting back on the freeway, as all northbound buses use the Southport Drive to I-405 N ramp, which has a meter queue that often extends much past the point where the bus can bypass the queue, requiring the bus to wait 10+ minutes sometimes just to get back on the freeway.
• 566 doesn’t run all the way to Auburn sometimes, serves the I-405 freeway station sometimes, but also only runs from Auburn to Renton, sometimes. This leads to confusion for both passengers and operators, major inconvenience for people who jump in and get on the wrong bus, and undoubtedly scares people who currently drive from taking the bus because they don’t think they can figure it out.
• 560 service from west of the airport is particularly long and circuitous, and does not have either speed or frequency improvements at peak, where the experience is the worst.
• Service to Overlake on the 566 and 567 is very inefficient, with a ton of recovery time at Bellevue TC southbound, which takes up valuable space in the transit center. Traffic between Overlake and Bellevue TC is very volatile, so this recovery time really is necessary to ensure timely service from Bellevue TC (where the bulk of passengers board).

While the opening of Stride BRT in four years will solve many (but not all) of these problems, four years is too long to wait for solving problems that keep people in their cars and punish people who choose to take the bus when it matters the most. I have a proposed restructure that addresses all of these issues, though it is almost certainly not revenue neutral.

Here is my proposed network map. Note the peak and off-peak layers, as the network has entirely different sets of routes for off-peak than peak.

Here are the list of peak routes:

• Route 546: Overlake to Bellevue (every 10 minutes)
• Route 561: SeaTac Airport, Renton, Kennydale Freeway Station, Bellevue (40 minutes)
• Route 562: Westwood Village, Burien, Renton, Kennydale Freeway Station, Bellevue (40 minutes)
• Route 563: The Landing, Renton, Newport Hills Freeway Station, Bellevue (20 minutes)
• Route 567: Kent, Bellevue (every 20 minutes when 568 is over capacity, timed with Sounder)
• Route 568: Auburn, Kent, Bellevue (every 20 minutes. Timed with Sounder at Kent when demand is low, halfway in-between Sounder trains when demand is high and 567 is running)

Off-peak routes:

• 560: Westwood Village, Burien, SeaTac Airport, Renton, both freeway stations, Bellevue (every 30 minutes)
• 566 Auburn, Kent, Renton, The Landing (every 60 minutes. North and south bound timed to meet with 560 for service to/from Bellevue, similar to today)

One note about Overlake service on the 546. It is meant to be dynamic, meaning that the operator (which would ideally be Metro) would start with the 4 or so buses required to maintain 10 minute headways under ideal conditions (which is very fast). As conditions worsen, buses may be added to the 546 if necessary from peak directional trips on routes 545 and 550 which would normally go to Bellevue Base. So traffic is bad on 520 and 405 and the buses in service on the 546 will not be able to maintain headways, 545 and 550 operators who finished their last peak runs and who are already near a terminus of the 546 will add their buses to the 546. 546 southbound will also dynamically switch from I-405 to 112th Ave NE when I-405 gets slower (Exactly like the 232 does all the time).

All restructures are tradeoffs, with winners and losers. Here are the big winners in this scenario:

• Kent and Auburn to Bellevue riders (and also, via Sounder, Sumner, Puyallup, Tacoma, and Lakewood to Bellevue riders), who see an improvement in both speed, frequency, and capacity on the 567/new 568 pair.
• Riders on the I-405 freeway stations going to Bellevue or Renton, who see both a frequency boost (from 30 to 20 minute headways), and a speed boost compared to the 560 today.
• Potential riders at Myers-Briggs P&R and the 128th street/SR 509 freeway station, who are getting service to Renton and Bellevue for the first time. This is due to the 560 and 562 using the 509 freeway, which is very fast in the reverse peak direction (which is really the peak direction for riders heading to Renton and Bellevue).

Here are overall winners, with some tradeoffs and sacrifices:

• Renton to Bellevue, who are no longer subject to traffic on SR 167, and are only subject to either traffic delays from Burien or from the airport (not both). Service is faster in Renton, the bus gets on the freeway faster, and the 10 minute service is real 10 minute service. Rather than piecing together 10 minute headways with both fast and slow buses, buses are all medium, serving one freeway station each. This slows down all buses from Renton a little bit, but they are slowed equally, while still giving the freeway stations more and faster service.
• Burien and Westwood Village to Renton and Bellevue riders, who get a much faster ride during peak, at the cost of headways going from 30 to 40 minutes. Though farther from frequent service, the quality of service is much better, and probably usually saves more than 10 minutes per trip.

Here are people whose experience is roughly a wash, or a little bit worse:

• Riders from Overlake in the evening, who now have to transfer and will very often get different buses from Bellevue depending on how bad traffic is.
• SeaTac Airport to Bellevue, who get reduced frequency (40 minute headways), but get a faster trip through Renton and on I-405 (due to serving one less freeway station).
• Riders from the Landing (north Renton) to Bellevue, who currently enjoy being the last stop (or near the last stop) before Bellevue currently. They keep a one-seat ride, but their frequency drops to 10 minutes and they have to backtrack to Bellevue TC first. Once on the freeway, the have the benefit of faster routing and only serving one freeway station on the way to Bellevue. They also have the benefit of getting first dibs on seats on the bus, and have the most reliable trip (as the fewer buses that serve the Landing begin at the Landing).

Finally, the losers in this scenario:

• Auburn to Renton and Kent to Renton (particularly to the Boeing plant) are the big losers, needing to take alternative service during peak. The best alternative for these passengers is usually to take Sounder and transfer to the F-Line.
• Passengers going between the two I-405 freeway stations. These riders can take Metro service like the 101, 167, and 342, but overall have much lower frequency. These riders are not affected off-peak.
• Riders along Ambaum Blvd near a 560 express stop, who enjoy current 560 service. They will require a transfer all the time, to the 120 (and later the RapidRide H-Line).
• Peak riders from the airport to Burien or Westwood Village, who will need to take infrequent route 180 to Burien, and then possibly another bus to get to their destination.

It is certain that this restructure would help many more people than it hurts, but it is likely more expensive to operate than the current network. It would probably require significant increase in spending in the East King County subarea, since there would be more bus service overall on I-405, and one new route that is entirely within the East King subarea. Other routes may be able to draw on funds from the South King subarea, and the 562 in particular has some speed improvements that may make it efficient enough to run with one fewer coach off-peak than the 560 does. Combined frequency from Renton could be adjusted to be 15 minutes instead of 10, and that might make it revenue neutral (and to be honest, I kind of prefer real 15 minute frequency to fake 10 minute frequency). But one thing is for sure: waiting until 2024 to improve ST Express bus service here that has numerous issues will hurt ridership once the BRT line is open, and will keep people in their cars for longer when traffic is at its worst. And that is unacceptable.

Metro has recently posted updated recommendations for their upcoming Renton-Kent-Auburn mobility bus restructure. I find these changes exciting as a former rider of some of these routes, especially considering my former weekly 3-seat evening trip from Kent to Federal Way. Though understandably this has gotten less attention than other, more significant restructures like the North Eastside restructure.

Overall, the restructure proposal is quite solid, and is a major upgrade for existing service in the area. However, I have my own take on the proposed changes and some different recommendations for some changes. First, here’s a link to my own map that shows only routes that are changed from the proposal or status quo, as well as entirely new routes that replace other routes in the proposal. Though note that I have not done any service hour math on these to see if it is revenue neutral (though I suspect that it might be), and I have not done any formal route planning work or training.

Routes 102, 148, and 906

Metro notes that Fairwood riders of route 102 to Seattle have a long and slow trip, and that a Sounder connection on route 906 would be faster. While that may be true, it seems to me that the slowness of route 102 is more of a problem of bus routing than an insurmountable obstacle. The fact that route 102 runs as an express for such a short portion of its route between Renton and downtown Seattle seems to be the real issue. If that weren’t the case, then it seems to me that riding a revised route 906 in two loops around Tukwila before transferring to Sounder would not be faster at all.

So my proposal is to move route 102 to run on I-5 south to I-405, and then exit at Rainier Ave S. in Renton, which brings it right to the doorstep of S. Renton P&R. I also propose not making routing changes to routes 148 and 906. Why I-405? It seems attractive because it has good direct access ramps to/from HOV (SB 405 HOV lane becomes a ramp to the left lane of I-5 north, and I-5 south has a left-hand ramp to I-405 north, which is easily accessible from the HOV lane). It also avoids local stops on MLK (which would still be served by the 101) and Rainier Ave S. in Renton (though that’s not so bad since it has BAT lanes). More importantly, route 102 could stay on I-5 until Seneca street in downtown Seattle, skipping SODO, and using the saved service hours to run to South Lake Union to increase the user base. It seems to me that before complaining that the 102 is just too slow for Fairwood, they could at least consider making it an actual express before they throw in the towel.

F-Line, New Route 110

I’ve proposed making the F-line faster for every trip, by making the connection to Southcenter more brief, and spending more time on Southcenter Blvd. It would require HOV or BAT lanes on Southcenter Blvd, which could be from 61st Ave to Interurban Ave eastbound, as well as SW Grady Way east of Interurban Ave westbound (which would probably require some widening or lane reconfiguration on the bridge), and a bus queue jump at the normally right-turn only lane at Interurban Ave. If it sounds like much, it’s not really much compared to real BRT, but these are the kinds of target changes that could make RapidRide F Line at least a little worthy of its name.

This means that RapidRide F line would skip Tukwila Sounder Station. This is because Sounder has only a handful of morning trips and a handful of evening trips (and a few reverse peak trips), and that doesn’t warrant all-day, frequent 7-day service unless it can be served reasonably on the way (which I don’t think it can). For replacement service, I don’t think there’s anything wrong with timed, targeted Sounder service (like there was before the F-line restructure in 2014). So my proposal is also to bring back the 110, restoring a timed Sounder connection to Boeing Renton and Kenworth. The bus would skip Renton TC since its ridership would come from Sounder.

Route 157

It’s curious that Metro recommends transferring route 102 riders to Sounder while not route 157 riders. Route 157 could be more directly routed and truncated at Tukwila Station without being circuitous. If that were to happen, then the very small number of trips (4 morning, 3 evening) could be expanded, perhaps enough to connect to every peak direction train. For efficiency, this route could through route with route 110, something that works because route 157 would drop passengers off who would get on the train, and route 110 would be meant to pick passengers up from the train. So operating both with one bus could be quite efficient, even more so than running trippers all the way to downtown Seattle.

Routes 165, 191

Another weird quirk of Metro’s proposal is the continuation of redundant bus service to downtown Seattle from Kent Station and places east (new route 162), while eliminating routes 158/159. And in fact, they are making this one-seat ride to Seattle slower by merging it with route 192.

My proposal is to not do route 162 at all, and replace routes 158, 159, 190, and 192 with two routes: a north-south route primarily on Military Road to connect to Angle Lake station (route 191), and an east-west route primarily on S. 260th, Reith road, and Meeker street, which is timed to connect to Sounder trains (route 165).

Route 191 would run like route 190 from Redondo Heights P&R, run on Military Road like route 190, and continue on Military Road (or could also take a short hop in I-5) until S. 200th Street, where it could end and connect to Angle Lake Station. This would be a hard sell because everyone’s trip would likely be slower, but in turn, route 191 could run very frequently or possibly even all day. And while Link would probably be slower than an I-5 express, it would be a consistent trip, and passengers would get first access to coveted rush-hour seats on the train. It would also greatly simplify access to other places like UW and the Airport.

Route 165 would bring new service to certain areas of Kent and Des Moines. It would start at Highline College where other buses have a layover space, and run on 16th Ave S (a major neighborhood corridor paralleling the even more major Pacific Highway corridor) until S. 260th Street, turning left and following the road as it becomes Reith Road and then W. Meeker Street. Then it follows the current 183 route, except following Lincoln Ave to James Street P&R, and dropping off at Kent Station timed to meet Sounder to Seattle. Routes 191 and 141 together (along with routes 166 and 183 both running east-west to the north and the south, respectively) would create a robust peak coverage grid, giving a large portion of local residents access to some form of rail service to Seattle with a connector bus.

Route 141

This is another Sounder connector route created to fill in some of the gap made by speeding up the F-Line. The route would run from Angle Lake Station, following a coverage route through SeaTac and Southcenter, connecting to Sounder trains at Tukwila Station. Service would partially duplicate route 156, so route 156 could be consolidated to run in both directions on Military Road and S. 164th Street, while rotue 141 could take the other branch on S. 170th Street. The route would through route with the current route 154 to Federal Center South in exactly the same was as my proposed route 157 would through route with proposed route 110. Additionally, the reason route 141 would start at Angle Lake Station even though very few if any riders would ride from that far to get to Sounder is that certain trips from route 191 could become route 141 (after a longer than usual layover, to ensure that it gets to Tukwila in time to make the train reliably), so it’s operationally efficient in this scenario to start all the way at Angle Lake Station.

This means that on the whole, instead of spending a ton of money on trippers from park and rides to Seattle, we instead have three routes (191, 141, and 154), all with different use cases and audiences, all being able to be run with a single bus.

Route 183

Metro’s proposal has route 183 run every half hour rather than every hour on Saturday, but still not run at all on Sunday. I propose instead running the bus every hour on both Saturday and Sunday, and expanding frequency later when more resources become available. Especially with route 166 being moved off of Meeker Street, and considering that the rest of route 183 serves a unique area with no other options, expanding baseline service to 7 days/week seems more important than expanding frequency at this point.

Route 184

I suggest extending service from the proposed route 184 (the south part of current route 180) into Lakeland Hills, currently only served at peak by Sounder connector Pierce Transit route 497. It is a opportunity created by separating off route 184 from route 180. Funding for service could be done by consolidating with route 497 and maybe working out an agreement with Pierce Transit. Marginally not a ton of additional ridership would be added, but in aggregate this route could be quite popular and bring much more neighborhood transit access to more of SE Auburn.

Things are getting busy on the Eastside. Light rail tracks and guideways are popping up around Bellevue and Redmond, in a preview of the future of transit on the Eastside. But with the June closure of the Montlake freeway station, SR 520 is about to get busy as well. And with Metro considering various options for truncating 520 bus service at UW Station to take advantage of our existing light rail network, the old life of transit on 520 as an Eastside-downtown workhorse may be finally coming to a close.

Some of this has been a long time coming. Ever since the plan to connect Redmond and Bellevue to Seattle with East Link was approved, the days of the frequent 545 route to downtown Seattle were numbered. Once planning for U-Link restructures began, planners belatedly realized the potential utility in removing buses from downtown in favor of a UW Link transfer, a plan that has (to date) nothing to show for it except evening and weekend 542 service (and even that was likely only done because of the closure of the Montlake freeway station).

There have been various discussions on how to use SR 520 for transit in the long term. Sound transit 2 envisioned a Bus Rapid Transit (BRT) network, which has apparently devolved into more frequency on the most popular bus routes on the bridge already. There have also naturally been discussions on building light rail on SR-520, becoming more relevant since WSDOT has kept the option of light rail on 520 in consideration in recent designs (realistically meaning that the rail could be build without needing to demolish and reconstruct the main spans of the bridge). Seattle Subway has had shifting opinions on this, at one point favoring a new light rail bridge from Kirkland to Sand Point, then coming around to proposing light rail on 520 and connecting the line to a real First Hill subway.

Though the possibility of light rail on SR 520 is exciting and would dramatically transform how commuters in the Eastside get to Seattle, it’s easy to forget how large of a project that is and how far off it would be. Sound Transit is pushing the limits of its building ability with Sound Transit 3, meaning that work on a 520 line would likely extend well past the 2041 timeframe of the last ST3 projects. Further complicating matters is that construction beyond this point is facing fierce competition from other priorities, such as Ballard-UW, extensions of the planned Ballard and West Seattle lines, and filling holes in our network in First Hill and Belltown. With light rail coming to I-90, many will view an additional line on 520 as redundant and unnecessary. With many (and ofter better) projects competing for resources, it’s hard to think that SR 520 will get light rail on any reasonable timeframe.

The good news? I actually don’t believe that light rail on the 520 bridge is necessary at this point, for a few reasons:

1. The bridge itself does not have a dedicated bus lane, but it does have the next best thing: HOV 3+ lanes. Convincing WSDOT to convert an HOV 2 lane to HOV 3 is like pulling teeth, but the 520 bridge has them today. Furthermore, all vehicles (including HOV 3+) are subject to a toll, which is an additional incentive to get people to take transit across the bridge, or to take I-90 instead if they must drive (both of which leave more room on the bridge for buses).
2. There are already two left-side freeway stations on SR-520 between I-405 and the bridge, where light rail would be should we decide to build it. These stations, as they stand today, are almost as high quality from a bus priority and access standpoint as it is possible to build. To access them, buses exit from an existing HOV 3+ lane into a dedicated bus-only lane that extends through both freeway stations all the way to the bridge!
3. There is direct HOV and bus access at 108th Ave NE, allowing buses from S. Kirkland and Bellevue (both future light rail connection points) to enter the freeway in the HOV 3+ lanes.
4. Starting this year, SDOT is pouring money into a redesigned Montlake boulevard/SR 520 interchange that has direct HOV 3 access and inline bus stops, that UW-bound buses can use to avoid the long queue of cars at the general purpose ramp.
5. SDOT is building a reversible direct access access ramp from the SR 520 HOV 3+ lanes to the I-5 express lanes, which will continue as a fifth express lane until Mercer Street, which will be a direct entrance/exit to South Lake Union. When ramp opens in 2023, it will be bus only to start (though it’s unclear why HOVs would be initially excluded), with HOV access coming later. This completely solves an issue that has been puzzling transit planners for years, who have been contemplating painfully circuitous routing to try to directly serve SLU while avoiding a slow downtown. The interim solution is a decent-but-not-ideal route 544, which gets off at the first downtown exit it can safely reach, and heads up a relatively non-congested Fairview avenue, avoiding Mercer Street like the plague.
6. There are lower-quality right-side freeway stations in Redmond at NE 40th and NE 51st streets, in the vicinity of Microsoft and East Link stations. Though sub-optimal, these stations are past the biggest bottleneck (namely the bridge), and have the benefit of an unusual right-side HOV 2 lane running from I-405 to just north of NE 51st street (with breaks in it to allow SOVs to access freeway exits).

While this may not be the 520 universe you might build from scratch, the fact that this is all already happening and will be fully complete by 2024 with no tax increases, ballot measures, and fights over tunnels is something to be taken full advantage of, rather than torn up to squeeze light rail into. To that end, I propose a comprehensive network of SR-520 buses that serve places all over the east side, combining to form a “virtual light rail line” across 520 that could take advantage of every piece of HOV and bus priority to be nearly as fast as an actual light rail line.

Here are the route maps for peak, off-peak, and downtown Seattle peak detail.

Here are some broad themes:

1. Though this can be nearly as fast as rail would be, it still would probably be somewhat unreliable (though this could be mitigated with some schedule padding at strategic locations). It also has much less capacity than light rail, which is stretched today by peak-level demand on the 541/542. This is primarily mitigated by high frequency service, where there are multiple branches (4 peak, 3 off-peak) that are all frequent on their own, overlapping on 520 creating very frequent service (2.5 minutes peak, 5 minutes off-peak).
2. Bellevue is getting express service to UW. This is significant because Sound Transit currently plans on having riders from Bellevue to UW take the blue line south through downtown Seattle (a hefty 29 minutes). Though they need to draw the line somewhere, I feel this puts Bellevue at the point of highest inconvenience, and many riders (especially those heading to places north on Link) will be turned off by needing to do a large amount of backtracking.
3. Sammamish would be well-connected to the regional transportation network, with nearly every destination being two seats or fewer away. North and south Sammamish service is now split at S. Sammamish P&R, where the north route takes the 520 bridge and the south route takes the I-90 bridge. There is no longer a dedicated Sammamish route, as now the Redmond and Issaquah routes simply continue to Sammamish at half frequency, eliminating the need to transfer. This makes Sammamish, which (not incorrectly, in my opinion) felt like a non-consideration in ST3, less of an afterthought by making some sensible tweaks to bus service that already terminates near Sammamish.
4. South Lake Union is a major destination, which is served during peak by the Redmond lines during (the Kirkland/Bellevue lines go to UW during peak, and all lines go to UW during off-peak). Since the ramp to the I-5 express lanes/Mercer street is reversible, only going one direction at a time, the configuration proposed has a loop through downtown (ideally with a 5-10 minute layover somewhere downtown for buffer time) that starts at SLU and goes counter-clockwise in the morning, and starts at downtown going clockwise in the evening. This ensures that all 520 bus trips (both SLU and UW) have a connection to Link (Westlake and UW, respectively).

Overall, this would require a big increase in service hours, so realistically this kind of thing might be tied to a 2024 ST4 vote, in which ST could fund all these lines. But for the time being, I’ve proposed a Metro/ST split that flips the agency of some existing lines:

• The 555/556 would be replaced by the 254, and Metro would operate this route.
• The 269 and all the peak Sammamish and Bear Creek Seattle expresses would be replaced by the 543 (peak), 546 (off-peak), and the extended half of 554 trips, all of which ST would operate.

Here are some details by route:

• 255: The same 255 from Kirkland to UW that is coming in March 2020. Peak headways: 10-12 minutes (slightly worse than 2020), midday/weekend: 15 minutes (same as 2020), evening: 30 minutes (same as 2020)
• 254: A new route similar to the 555/556 that replaces both, and provides all-day service from UW to S. Kirkland, Bellevue TC, Eastgate and Bellevue College. S. Kirkland P&R is added because Sound Transit operates the 555/556 along 112th Ave NE instead of I-405, believing it to be faster, so adding a stop at S. Kirkland to this route is relatively non-disruptive (because it’s already on the road to S. Kirkland P&R anyway). This route also continues to Bellevue College and Eastgate, since this idea opens up the possibility of giving Bellevue College & Eastgate a fast ride to Bellevue TC all-day, and I don’t believe this route needs to continue to Issaquah (since the 554 will be more frequent by this point). My preferred routing has it take NE 6th to I-90 (utilizing the express toll lanes available in 2024), but this requires using the right-side exits to switch between 405 and 90, so routing along Bellevue Way or the 271 route (with no stops between BTC and BC) may be more sensible. Truncating the 271 in Medina (ideally at one of the freeway stations, and not cross the bridge to Seattle) could partially pay for this route. Peak headways are 10-12 minutes, midday/weekend: 15 minutes, evening: 30 minutes.
• 544: A variant of the 544 coming in 2020, except more resembling the 545. This route would replace the 545, and would serve downtown Redmond and Redmond TC (rather than Overlake P&R). There would be no stop at S. Kirkland P&R (but an added route 254 would ease transfers to this route from S.Kirkland). It would serve SLU using the previously described loop (a more convenient route than route 544 in 2020). Unlike the 545, this route won’t go to Bear Creek, leaving that to the 543 (which offers a faster and more direct route to Seattle than the 545 does). Peak headways: 10-12 minutes (slightly better than 2020). No service off-peak, as that is offered by the 546.
• 543: A different variant of the coming 544, except skipping downtown Redmond to serve Bear Creek and Sammamish. Meant to complement the 544,this route takes a direct route between Sammamish, Bear Creek, and SLU. Residents of Sammamish who want to take transit to Seattle or Overlake, but struggles with the weird peak-oriented service that takes them down to I-90, would find this route easier to use. Peak headways: 10-12 minutes, no off-peak service (route 546 is available off-peak)
• 546: An off-peak-only route that combines the 543 and 544, and goes to UW instead of SLU. Service follows the path of the 542 in UW, the 545 in Redmond, and half of the trips continue to Sammamish. Midday/weekend headways: 15 minutes, weekend: 30 minutes.
• 554: This route will be largely unchanged, presuming it exists in some form in the East Link restructure. It will likely terminate either at Mercer Island Station or South Bellevue Station (in this case, likely also continuing to Bellevue TC, but could sensibly terminate at S. Bellevue if route 254 is introduced). The important tweak to this route would be extending half of the trips from Issaquah Highlands P&R to South Sammamish P&R, providing the southern portion of Sammamish access to Seattle and Bellevue via I-90 (whereas the northern part accesses Seattle and Bellevue via 520 and the Redmond routes). It seems like such an opportunity that frequent 554 service is coming nearly to Sammamish, and then stops (except for a few late night trips on the 554 heading back to the base through Sammamish). Headways would likely be comparable to the 543/546, but depend on what the headways for the restructured 554 will be.

With that, I think this transit network on 520 will provide service comparable to light rail on the bridge, without:

1. Absolutely needing to transfer to another bus route to get where you want to go on the eastside.
2. Spending a lot of money, years, and political capital on building rail when there is already top-tier bus stations and freeway access either built already or under construction.
3. Focusing Link construction away from places that really need Link light rail service to get much better (like UW-Ballard, Metro 8, or a number of others), and better bus service simply won’t do it.

Let me know what you all think in the comments!