One point on demand bears mentioning here: I know a lot of these trains won't make money, but it's a lot easier to defend a train with 75% CR than one with 40% CR. It's also easier to defend a full train (and it might be worth trying to get Amtrak to start giving some more diverse capacity stats for boarding/alighting at different cities on specific trains). It is at least somewhat harder to kill off a packed train than it is to kill off an empty one...and let's not forget that every so often, the process of killing off the empty train can cause a fit for the full ones as well.
So I'd have to wonder what the load factors/CR would be for a train twice as long as the current one operating south of Atlanta. If the present situation is that the train is well under half full through that section, doubling the length is going to make the financials there just awful. I mean, I'm thinking about something I read when the Crescent derailed a bit ago. 60 passengers leaving New Orleans. The train probably has a capacity of 200 to 250; I'd guess you're looking at about half full when it reaches Atlanta. "two more coaches, three more sleepers, and an extra diner" is going to come close to doubling the train's capacity, so you'd be looking at probably generally a load factor of between 10 and 20% into New Orleans, with a full dining car staff, three or four sleeper attendents and possibly a few more people chilling in the lounge and burning money. That has to be the most inefficient use of personnel and equipment ever. (Of course, you can always cut off a lot of the train in Atlanta, but then you have the staff chilling in Atlanta and burning money.)
So I really think that adding a second frequency, whether it's a day train or an overnighter, is a more efficient use of resources than blindly adding capacity onto the existing service.