My only real disagreement with the analysis is with the skepticism about the private sector investment. It's one of the more neglected bits, but I can't help but wonder if on some of the Eastern LD trains (namely the Silver Meteor, Auto Train, and Lake Shore Limited) we couldn't see someone propose something along the lines of the Caledonian Sleeper. The Hoosier State situation is something of a "perfect storm" (it's an insanely complicated route, it doesn't run daily, Indiana didn't have half a clue what they were doing, and Amtrak hasn't exactly been a helpful partner to put it lightly).
Put another way, through the end of August 2015 (for FY15, so 11 months) Amtrak posts a "pre-asset allocation loss" of $38.1m on the Star, $30.3m on the Meteor, and $11.0m on the Palmetto. In theory, that's $68.4m in losses on the books for a Meteor/Star contract or $79.4m for a "full service" contract. On the basis of Boardman's figures from 2012 (or thereabouts) the combined losses here come to roughly $1m/yr in direct expenses.
So, what happens if someone comes in and says "Give me a RRIF loan for new, expanded sets of equipment and I'll bid the cost of paying off that loan to Amtrak"? Eight 18-car sets of equipment would run about $360m (assuming $2.5m/car); you could add 15% spares [1] and you're still looking at "only" $415m for the cars and $70-96m for locomotives [2] [3]. I'm presuming they expect to cover RRIF interest (at something like 3-4%) out of operations...which in such a scenario would be comparatively unconstrained by equipment shortages (e.g. the trains could run with 6-10 sleeping cars as demand dictated).
[1] Probably a reasonable proportion...keep a coach, sleeper, and food service car at each end as spares (6 spares, maybe 8) and be willing to vary train sizes by a bit here and there and you probably wouldn't need to keep 21-22 spare cars in reserve) and run your annual inspections between 9/15 and 10/31 and simply plan to run the trains short for that slow period (cut from 18 cars to 14 or something like that...or do the sleepers then and the coaches in Jan/Feb). Or just effectively buy 9 sets plus a small set of spares for one end of the route, which would have the same effect.
[2] The locomotive numbers vary a bit since I don't recall whether to presume $5m or $6m for a diesel locomotive, and whether to presume 8 sets of locomotives or 7 since the diesels aren't needed north of WAS.
[3] Why "only"? Well, $415m over the course of a 15-year contract comes to $27.6m/yr and $511m comes out to $34.1m/yr. Compared with present accounting losses, Amtrak notionally shelling out $34m/yr rather than booking $68m/yr in losses would...well, explaining turning down that offer to Congress would get very interesting very quickly.