Sooooo, there are some studies -- I do not have the citations on hand -- which show that people judge trip time by a 10th percentile or 5th percentile rubric. So with the 5th percentile rubric, if 1% of the trips take 12 hours, people blow it off as a one-time event -- but if 5% of the trips take 12 hours, they treat it as taking 12 hours, even if most of the trips take 4 hours.
This means that if there's variability in trip time -- like there is with Amtrak due to corrupt freight rail companies causing delays -- people mentally judge the trip time by the worst 5% of trips. All of you rail travellers can probably confirm that this is how you think, though there's probably some variation in whether it's worst-5% or worst-10%.
This means that on-time performance is, in many ways, far more important than scheduled trip time, in terms of perceived trip time, and therefore customer acceptance and customer demand.
I believe that the purported distinction between "corridor" trains and "long-distance" trains is essentially bogus, and I would classify the NY-Pittsburgh-Chicago corridor, the NY-Albany-Buffalo-Chicago corridor, the Chicago-Denver corridor, the NY-Miami corridor, and the NY-New Orleans corridor, and the Chicago-Minneapolis-Fargo corridor, as corridors.
The Salt Lake to Bay Area route probably doesn't qualify as a corridor, and the same with the Eugene to Sacramento route, and the same with the cross-Rockies portion of the Empire Builder and the Raton Pass section of the Southwest Chief, and the desolate sections of the Sunset Limited between San Antonio and Tucson. Call these outliers "system connectivity routes", perhaps. These all have the characteristic that there's a severe absence of intermediate population centers. A corridor, in my opinion, has decent-sized cities located less than two hours apart.
But most of the so-called long-distance routes are corridors. The problems with these corridors right now are that (a) they are not running on time, and (b) they are not fast enough even when running on time (even between intermediate cities -- we're not talking end-to-end), so they are not working as corridors right now. They also (c) don't have enough frequencies per day, but I think that is largely due to (a) and (b). I think the top priority for "corridor development" should be the Lake Shore Limited because it interconnects the Northeastern network with the Chicago-centered network, so it has the maximum benefit in terms of the synergies caused by network effects.