I don't see how it could be an ADA issue; the cars themselves don't allow wheelchair through travel down the hallways & aisle (they should, but they don't) and the car next to the dining car is oriented correctly, and the car next to the cafe car is oriented correctly.
It can't be a signalling/shunting issue; the trains are long enough.
It can't be a crash-safety end-car issue; baggage cars are on the end and they're the newest.
I don't see how it could be a PTC issue; this would imply
gross incompetence in the PTC design, which should be able to handle any train consist whatsoever given suitable programming. (If it doesn't, my services as a programmer are available; I can redesign the system for you. I am very, very good at this stuff.)
It can't relate to boarding safety at Penn Station, because the platforms are the problem there, and that affects all trains.
There's already a long, time-consuming sequence of engine off / engine on / brake tests etc. after the join or separation, so it can't have anything to do with any of that. That does take time, but it's largely consumed within the recovery time and time for meets with the other section, so eliminating it probably wouldn't actually save any time, unless the train could be made to run on time on the *rest* of its route, which sounds impossible in the medium term. Furthermore, the transferring passengers dragging all their luggage around would eat up substantial amounts of time, "cross platform" or not. I just don't see a time savings here.
I don't see what other regulations there could possibly be which would apply. There really are no other legal restrictions on train consists.
If "Part of the standard is grounded in law", I'm strongly suspicious that there's nonsense and **** afoot and that someone needs to get a bad regulation revoked. The FRA does have a history of bad regulations.
It can't relate to frequency of inspections or to staffing if the "nonconforming" issue is on the New York section... the Boston section travels for more hours when it goes to Chicago.
And if it does relate to staffing, the staff can always be changed at Albany.
Seriously, if there are worries about "conforming", Amtrak's priority should be reconfiguring Hudson station or at least changing procedures there, since it has an inherently unsafe boarding situation involving passengers crossing active tracks. There are also several SEPTA stations on the NEC north of Philadelphia which are basically invitations to liability lawsuits.
Without know that portion and what it entails, how can you really say if that would result in less efficient utilization let alone say that it would cost more. In the grand scheme, it may indeed be more efficient to run a stub section with a cross platform transfer than altering a bunch of other trains on a routine basis.
Because it would result in less efficient utilization. This is something which is measurable. I was assuming a sleeper on the BOS-ALB section, and you get less efficient utilization.
If they dropped the BOS-ALB sleeper, then sure, they might have slightly more efficient utilization. At the expense of ridership and revenue. Good way to drive the company down the drain.
It's also very dangerous to run the BOS-ALB section as a completely independent train with no join/split. It triggers the PRIIA rules for state-supported trains, and I don't think Massachusetts will be happy about that. They might decide to contract with someone other than Amtrak.