We seem to be dodging the question here. Is there anywhere on Amtrak where a DMU offers something above-and-beyond what the current fleet allows...for Amtrak operating practice, not commuter rail or somebody's favorite Euro reference spec taken in a vacuum?
-- Amtrak doesn't have VIA Rail's federally mandated routes to tiny fishing villages. Likewise, there are quite a few European DMU routes that split/combine trains for a low-demand branchline only because the low-demand branchline(s) have government-mandated service regardless of their ridership. That is not a consideration in the U.S.
-- Amtrak doesn't have any routes like VIA's suspended Malahat that are isolated from the national rail network on a large island.
-- Amtrak has very few routes that split/combine trains (one of the main applications where DMU's in Euro intercity are utilized). There are very few routes proposed or desired where split/combine would provide any tangible benefit that running additional frequencies on a forked route wouldn't solve more easily and robustly.
-- A state that cannot afford to pay for additional frequencies on a forked route will most certainly not be able to afford the going rate for an ownership share in a PRIAA-spec DMU fleet. A state that cannot find enough demand to pay for the operating cost of additional frequencies on a forked route will most certainly not be able to afford the operating costs of a split/combine route. If either of these conditions are in-play that state will quite likely also not be able to afford the capital costs of constructing the very branch they want to fork. Caveat emptor, Northern Maine.
-- Amtrak doesn't have many push-pull routes period. The push-pull routes it does have tend to be some of their busiest and most car-hungry routes, not applicable to DMU's. Most new push-pull routes officially proposed likewise overwhelmingly trend to higher-demand, equipment-hungry, multiple daily frequency service.
-- The most oft-cited prototype for DMU's on Amtrak--the one that used to run DMU's, the Springfield Shuttle--is not a prototype route for new DMU's on Amtrak because it only exists as a substitute for a long-missing commuter rail line. A commuter rail line that is at long last going online in another 18 months and which will largely supplant the Shuttle. Proposals for repurposing the Shuttle are either short-term, like pokes north to Greenfield or Brattleboro as interim incubator for eventual Knowledge Corridor commuter rail, or much larger in scope, like New Haven-Boston which will fill >4 cars and cease running as push-pull. Lines that should be commuter rail lines are not an invention for DMU's on Amtrak. The Shuttle runs small because it's such a very, very poor substitute for real commuter rail frequencies.
-- Perennially underfunded trains that run short because their frequencies are artificially crippled by state-level budget stinginess are not DMU prototypes. The Hoosier State wouldn't be running short if it weren't a perennial political football. The Heartland Flyer wouldn't be running short if the various high-demand expansion proposals for roping in a third state had more political support. States that are more ideologically reluctant to fund intercity service are the least likely to want to invest in specialty rolling stock. Didn't we find that out the hard way with the Wisconsin Talgo debacle?
-- There is very little fleet scale to be had in a DMU purchase unless routes can be identified that constitute more than one-off exceptions or misfits. The Cascades is the system's 8th busiest route, with 7 dedicated trainsets of 12-13 cars each; that's the only reason it can support a dedicated Talgo fleet. Are there enough prospective DMU routes in the entire country to support a PRIAA-spec order large enough to 1) net a decent unit price from a manufacturer, 2) net decent parts/maintenance cost scale, 3) group enough of the fleet around common maintenance bases around the system? The PRIAA coach and locomotive orders the states are buying and will be buying cluster around big regional nodes: California, Chicago hub, East Coast and New York State. Scale is everything in those orders, and the states that feed out of a large hub have many more options for initiating or expanding state-sponsored services at lower cost than those are isolated from a large equipment hub. Especially now that equipment is being mass-procured to ease longstanding shortages. If there is only a need for 20 DMU's to run widely scattered niche services in Maine or the Plains, operating costs for those individual states soar in a way the low margins on those services can't close. Where does the ordering scale come from without either a really big and car-hungry route like the Cascades or a big multi-route hub? If there isn't an operationally and economically ideal DMU application that is uniquely large, or concentration of DMU route clustering that is dense from a particular hub...it is very hard to find scale that can justify a unique equipment purchase. We have to do better than coming up with discrete and widely-scattered exceptions that all trend small in ridership and consist size. That's not going to cut it.
-- The PRIAA legislation's actual effects impose more fleet uniformity than fleet fragmentation. Don't cite the Piedmont as some sort of "anything goes" model for custom rolling stock just to dare to be different. Everything from rolling stock maintenance to track maintenance is wholly-outsourced by NCDOT to an outside private company, with Amtrak supplying the crews and giving it *basic* national conformity for ticketing and customer service. The Piedmont isn't functionally much different from INDOT more or less opting the Hoosier State out of the AMTK system for the Iowa Pacific outsource. NCDOT just did it as part of the plan when it began that route discontiguous from any Amtrak hab; INDOT's move was one of manufactured political crisis. Don't cite the Talgos as a model for custom rolling stock; the Wisconsin debacle ended that experiment in openness and bottled up availability of those custom rolling stock choices to the Cascades giants of the system and the states willing (for better or worse judgment) to open a full-time factory. The hurdles are massive for going it alone, and not conducive to going way way off-script with rolling stock. You're not going to get new DMU's by hiring Iowa Pacific; you're going to get very old and conservatively-managed conventional coaches. While I would not put it past NNEPRA to say something monumentally stupid about coveting things it has no means or mechanism to pay for, they are not getting Nippon-Sharyos to run a microscopic Downeaster fork to Lewiston any more than they're getting their own Talgo factory.
So what are the realistic possibilities? I'm having trouble thinking of any, since most of the route candidates that fit the profile go at the most unrealistic small end of the scale and are the exact inverse of what'll float a PRIAA equipment buy or a private outsource deal.