To expound on railroad consisting and the dangers it has when considered for passenger service:

One of the flaws that seems to emerge every time some sort of (prototype) trainset is put into service in the U.S. is that variable passenger loads mean it's hard to have a fixed consist and still have a serviceable business model. Basically you end up needing extra cars to service peak loads that you don't use for anything the rest of the time because they're tied to the particular type of equipment you're using and not useful for general train service.

Acela has a fairly well-defined utilization (because it's only business and first class, not coach) and so *may* avoid this problem. New York recently paid Super Steel to rebuild some old Turboliner sets which Amtrak may put back into service any time now, and there are Spanish Talgo sets operating between Portland and Seattle (maybe Portland and Vancouver) but other than that, nothing really, and those examples only offer a modicum of success, nothing resounding.