My family moved from Waterford, CT, one of the four towns that still has NEC grade crossings, to San Diego, many years ago. My parents drove two cars the whole way. I think between having only one potential driver per car, one of the cars being about 11 years old at the time, and the large amount of driving involved in that trip, an Auto Train might have been desireable. In fact, for cases like that, an Auto Train might be worth having even if it was more expensive than simply constructing more highway.
One Auto Train trip a week probably would not have worked well in that case; we left Waterford shortly after the movers did, and there was a strong desire to get to San Diego early enough that the moving truck could be unloaded into our new home instead of into a storage facility. (Other than riding the Northeast Corridor on several occasions without getting off at New London, I've never been back to that area since that day.)
However, with the current daily Auto Train being described as the longest passenger train in the world, I have to suspect that even if some new route was less popular, daily trips ought to be feasible.
I don't remember exactly where it's at but freight crews curse "Benson Hill" on the Sunset Route as being one of the toughest hills to climb. I remember a big hill west of El Paso and one east of Benson. Anyone know where it is and what grade it is?
One has to remember that the management of Amtrak is appointed by the President of the United States. That policy of state funded services as the only new services is probably actually a reflection of the desires of our current president. We're getting a new president in less than half the time it will probably take to build new Amtrak rolling stock once Congress approves it, so I hardly think the current president's policies should influence our thinking about where we want Amtrak to be five years from now.
I don't think the distance between the terminals is really the issue in the Auto Train catching up after a major delay. The real issue is how quickly the equipment is turned around, and how many hours the equipment normally sits idle while waiting for the next trip. If anAuto Train trainset spends X hours traveling for each trip, and it takes Y hours to load/unload/etc the train for each trip, and then it normally has Z hours of slack time when it's sitting in the yard and none of the railroad workers are actively doing anything with it, then if you double X and double the number of trainsets, you'll actually have an easier time recovering, because you haven't doubled Y, and so you can increase Z for the new train by whatever Y happens to be. And Z is the thing that you can borrow time from when you're recovering, potentially reducing Z to 0 during the period you're catching up.
I don't think Manhattan is the right place to put a terminal, though. Perhaps a train from New Jersey to Florida would make sense, and perhaps a New Jersey to Los Angeles train would make sense. Maybe there could also be a terminal somewhere around Denver, and at New Jersey, some of the autoracks could be designated for cars going to Denver, others for cars going to Los Angeles, and when the train reached Denver, the autoracks going to Denver could be uncoupled, and the Los Angeles section could be continuing along at 60 MPH or 79 MPH before the Denver autoracks were unloaded.
There's no inherent reason why there couldn't be a single-level Auto Train on some new route, although on a popular route, the train gets long enough that fewer, taller cars certainly becomes advantageous.
Do the Auto Train autoracks use slackless couplers? Also, how heavy is the typical Auto Train autorack compared to the typical maximum weight freight car?
