By the same token (and assuming the ability to do so...there are some cases where additions simply don't work), if Amtrak notices a train approaching, say, 95% of capacity a ways out they can also add a car if available.
In this context, there are definitely going to be various routes and subsets of routes that would be well-served by varying capacity (say, adding a coach on a given route for weekend-only use). Bear in mind that in the "old days" it wouldn't be implausible for an extra section to operate on certain days of the week in line with demand, or for cars to be added/dropped based on the same.
Of course, there would still have been heavyweights to throw around in the 50s and spare cars from dropped trains to use in the 60s. Regrettably, Amtrak had to ditch the Heritage fleet over the dump tank situation, which has likely cost us that ability (since at least in theory, a bunch of Heritage cars could have been pressed in during peak seasons)...but I digress.
I don't see these cuts as a horrible sign...in a lot of cases, based on raw ridership numbers, the cuts make some sense. I don't agree with how far the cuts went, particularly on the Coast Starlight (dropping a sleeper and a coach makes sense...especially if there's a plan to revamp the arcade space into something useful; two coaches, a sleeper, a transdorm, and the PPC? Not so much.). One thing to remember is that as far as I can tell, there's an increasing divergence between sleeper demand (which seems to be edging upwards where OTP hasn't totally disintegrated, and holding its own better than coach ridership anywhere OTP has collapsed) and coach demand.
Also...it looks like there was a plan to cut the CONO (based on other notes, possibly something involving dropping the Dorm [Paging Iowa Pacific...capacity needed on train 58/59...] and/or a coach), but some minimum axle requirement showed up in the files.
