Well, the $64m question is what the basis of the subsidy is. Bear in mind that (conceptually, at least) Amtrak is keeping 10% of the subsidy for overhead, Arrow, stations, etc., so the new operator doesn't have to have their own systems...
Remember the "Boardman chart" from a few years ago and how it showed far lower losses than the September (end-of-FY) Monthly Performance Reports tend to for all the LD trains. If the MPR figure is used then there is absolutely no reason that an operator couldn't operate at a slightly lower loss figure...especially if more equipment is brought in (e.g. an operator running a Meteor with 4-6 sleepers would probably be in the black vis-a-vis Amtrak running it with only three). A contract of 4-8 years at 90% of the MPR-indicated losses but funding the Boardman chart losses would probably give the operator enough room to mostly finance a new batch of equipment. A handy example is that the Boardman chart showed the Meteor roughly breaking even while the September 2015 MPR showed a year-end loss of about $31.9m for the Meteor. If the out-of-pocket losses for the Meteor are, say, $5m/yr (I'm presuming there are some indirect costs you couldn't avoid) but the subsidy the operator would get is around $28.5m, that provides a net of $23.5m/yr (or about $90m over four years/$180m over eight years). Over an eight-year contract that's actually enough to finance somewhere around 72 cars (or perhaps 64 cars and 4 locomotives)...basically enough to fully re-equip a much larger version of the train.
There are other places to improve performance, such as a fuller bar being available, actually stocking the cordials in some reasonable amount (which should have enough of a shelf life so as not to be worried about spoilage from trip to trip), improving turnaround time in the diners, improving cafe selections (like we're seeing on the Cardinal, etc. now), and getting the OBS to nudge pax towards a pre-dinner or post-dinner drink like is done on the Canadian.
Put plainly, the main place for improvement is in bumping up volume...given the prices of some of the trains (particularly in the East) and the fact that they'll still get close to selling out quite often, even if you knocked average sleeper fares down by about 10% but pushed to increase volume by 40-50% without a major shift to demand patterns you'd probably come out ahead.