"In fact, I don't believe there are any genuinely good routes for sleepers west of Denver."
Well, neroden, that's a broad statement. I didn't catch what your criteria for judgement are, but I sure hope your outlook isn't widespread.
Perhaps "genuinely good" is a vague and inaccurate phrase. What I meant was *commercially* solid, or financially promising. I agree that there are other reasonable criteria which one can use to judge routes.
I believe that routes west of Denver are going to require operational subsidies for the forseeable future, and that expansion of service would require enlarged subsidies. I think that there's a fairly low ridership demand level for any given ticket prices. It is probably a good idea for governments to support them -- heck, I take them, and I'd miss them if they were gone.
By contrast, I believe that there are a number of routes *east* of Denver where improved & expanded sleeper service would pay for itself over a reasonable number of years. There seems to be a much higher ridership demand level for any given ticket price. There are simply a lot more markets; extend any "corridor" route a little bit and you start finding lots of city pairs which are single-overnight sleeper markets. The vast empty spaces in the west mean that this doesn't happen there; extend a corridor route a few more hours and you're often in the middle of nowhere.
Given the continued difficulty getting government funding, I think expansion of these promising eastern routes should be prioritized over the western routes. Does that make my view clear?
However, *historically*, Amtrak has repeatedly slashed promising eastern routes while retaining expensive-to-operate western routes. The losses of the Silver Palm, Three Rivers, Toledo-Detroit train, and daily Cardinal service come to mind immediately, and the eastern end of the Sunset Limited may qualify as well; while the elimination of upstate NY - Chicago service on A-Day is the most extreme example of this sort of dumb mentality. I'm worried that this sort of poor decision will happen again.
Without sleepers, almost no route west of Denver would be bearable. I certainly wouldn't have convinced my family to take the CZ to Calfornia and connect with the CS northbound to Vancouver last Spring if I couldn't offer them cozy sleepers to rest the night away in peace. Western intercity routes are so long that few of them can be done within waking hours. And offering enough sleepers at a reasonable fare seems much less costly than rebuilding trains and tracks to HSR standards, doesn't it?
Well, no, that's the thing, I don't think it is "much less costly". I guess it depends what you mean by "HSR standards"; sure, 220 mph is extremely expensive (justified for SF-LA due to the nature of that route, though). Rebuilding and maintaining tracks for 90 mph or 110 mph, however, looks like it costs a lot, but there is a resulting boost in ridership, boost in revenues, and reduction in operating costs (the employees are working fewer hours per train-mile), which ends up paying for it, if you've prioritized correctly. (Examples of "not prioritizing correctly" include Illinois, which has 110 mph running in the middle of the Chicago-St. Louis line, but with tortuously slow running on either end.)
If I lived in the East, amid dozens of destination cities located five to ten hours away, maybe I too would be calling for speed and more of it. But out here, the lay of the land leads to a different conclusion, pardner.
It is very different in the east and in the west. We agree on that!
the East coast gets the marginal results it does from having a sufficiently large population base and consistently dense urbanization that you can find sufficient outliers.
This is pretty much my point, though I think there are so many that, statistically, they can't be called outliers any more. (My mom taught statistics. Outlier is a technical term.) "A sufficiently large minority group". (Or, to use the statistical term, "a sufficiently large cluster".)
The point was also made that a few city pairs lie in the sweet spot for sleepers so that they could appeal to time-sensitive people (if the trains run on time). However, this is largely a lucky accident. In the East, there are so many sizeable cities that you bump into a few of these city pairs when designing a route. In the West, it's almost impossible to design a route to have even *one* such city pair. This is a very similar phenomenon: "a sufficiently large minority".
Obviously there will always be a certain segment of the population willing to take sleeper trains. However, most of the population isn't comprised of railfans, and trying to convince them to take a sleeper which is slower than a plane, and smaller than a hotel room, (yet hardly cheaper usually) is a very tall order.
I'm sure there are city pairs where it could work. For example, take the Coast Starlight....
My entire point which I keep banging on about repeatedly is that it's relatively easy to find these city pairs in the east, and very hard to find them in the west. So please don't take the Coast Starlight as an example; for an example, please take the Lake Shore Limited, the Crescent, the Silver Star, or even the Cardinal!
I think it's important to differentiate the sleeper trains between the ones where scenery, service, and food are the big selling point, and the ones where time and location are the big selling points. On the "land cruise" (food, scenery and service) routes, you probably have a significant market segment who's more interested in the train ride itself, so maybe re-market the service as just that: a tourism line and such.
And on this note, you read people raving about the scenery on all four of the transcons, and the Coast Starlight, and sometimes the Cardinal. You don't hear much about the scenery on the Lake Shore Limited; sometimes about the Hudson River, but I think I'm the only one who likes the scenery in Gary, Indiana.

And you almost never hear people talking about the scenery on Silver Service, Crescent, CONO, Texas Eagle, Auto Train (which runs almost entirely at night), or the Capitol Limited (ditto).
It's probably not a coincidence that the trains with big "scenery" business are more seasonal than the ones without it. It's *certainly* not a coincidence that they do worse financially.