STRACNET clearance are there for oversized DOD cargos so regular gauntlet tracks need to be further offset. The only route that can really handle longer trains than posted would be the City of Everywhere thru Wyoming. That IMO just says it is Superliner style trains all routes west.
If enough Superliner cars were now available now the CZ, Starlight, and EB could probably sell 15 Superliner passenger car trains. Probably also the Eagle as it had 10 one day this past week. The SWC and CNO probably not. The Sunset needs daily service to tell but once west of SAS then it with the Eagle probably yes.
I think this claim is dubious. On the one hand, I suspect the Eagle
might be able to do so up on CHI-STL, where it shares the route with other corridor trains (and where, IIRC, a coach was cut out regularly a decade or two back). Further south, I'm not so sure.
As to the others - the Zephyr, in particular, is very seasonal. You could probably fill the train up close to 15 cars in summer, but you'd be back down to about six or seven in the winter. I did an analysis like a decade ago on here (good luck finding it with the search function) where I found that the Zephyr's summer ridership was like 2.5x the winter ridership. The Builder isn't as dramatic in terms of variation (there's a bit more "local utility" along the Hi-Line due to the lack of an interstate up there), but it still varies a lot. This is probably a good time to point out that the split fleet problem means that you can't move those cars to e.g. the Florida trains (where there was basically
no variation - all of the variations in ridership on those trains plus the Crescent were down to service disruptions, such as the Crescent being truncated to Atlanta for part of a few weeks due to track work).
It's worth noting that, per some data from the Washington State DOT rail plan back in 2013 [1], the Builder's ridership was at a substantial historical high in 2007 (at 554k riders). The Coast Starlight actually ran somewhat higher back in the 1980s/90s (this is probably why it was picked for the PPC project - it was touching 600k/yr repeatedly). In both cases, while equipment shortages haven't helped, the main culprit behind ridership declines have been reliability breakdowns which take
years to recover from.
Worth noting, too, is the fact that the Starlight hasn't run with more than three sleepers, a transdorm, the three service cars (PPC, diner, SSL), and a somewhat-variable number of coaches plus a baggage car in a long,
long time. In the meantime, I think the Builder had something like three sleepers (two SEA, one PDX), four coaches (two/two I think?), two amenity cars (SSL/diner), and a baggage car (one of the coaches was a coach-bag), plus an additional coach CHI-MSP. The point of all of this is that no Superliner train has run longer than about 11 cars, and when doing so you could accommodate 500-600k riders/yr.
Pushing that out to 15 cars would (1) either seriously stress F&B capacity or (2) require an additional F&B car, and would probably imply ridership up in the ballpark of 650-800k riders per year, depending on questions of whether you add coaches vs sleepers as well as seat turnover and where your "pressure points" (that is, segments where ridership peaks out, thereby squeezing out other through-riders - the Zephyr has two (west of Reno and west of Denver) but also has an infamous "ridership crater" between Reno and SLC) end up being. With the Builder, I think there's a strong case that if you're climbing up towards 15 cars you should be having a
very serious talk about making 7/8 and 27/28 into separate trains, though that's a separate discussion in some respects.
[1]
https://www.aawa.us/site/assets/files/7322/2014_wsdot_state_rail_plan_2013-2035.pdf
