You can forget about Northborough/Clinton Regional Rail ever being a thing if that's the final configuration. If we're eventually going to have overlapping all-local :30 Framingham short-turns layered with :30 semi-expressed Worcester turns, the Fitchburg Secondary makes for an extremely attractive direct extension of those Framingham all-local turns because of the large park-and-ride potential at Pike/MA 9, 495, and 290 plus hitting Framingham State U., dense downtown Northborough @ US 20, and large TOD potential at the 290 stop. Northborough would fit snugly into a 1-hour EMU'd local schedule, with a significant percentage of MetroWest's future transit shares driven via that extension of those schedules. So 100-year solutions for the grade crossing eliminations really shouldn't ever preclude 25-year solutions for MetroWest transit expansion. Especially when the envisioned Regional Rail service layer cake on the inner Worcester Line tees up such a branching so elegantly. The viaduct, and the general trajectory from that easterly replacement Framingham Station on the viaduct, is going to need to permit revenue movements north on Framingham Jct. Just leaving the ground-level wye (which CSX still needs to use multiple times a day to reverse the running direction of its locomotives) for freight-only with no viaduct or Framingham Station access won't cut it, without cutting off nose to spite face. That consideration is going to ding the budget for the crossing solutions in some significant way, even if you have multiple options for designing it.
Nevins Yard can also only get so much bigger before it gets inefficient. Arrivals/departure staging and sorting functions are currently fileted between Nevins and North Yard. Nevins takes generally long/intact strings of cars, while North does short strings of cars on a lot more tracks because stuff is constantly being pried apart and recombined. Google Maps illustrates this with the way the cars are arranged in each yard on satellite view. You can't simply take the total car counts of the two yards and combine them into one denser space, because one set of functions (the switching) is so much more space- and track-intensive than staging completed manifest trains and it starts presenting escalating safety issues the closer you pack those functions together at one time. You'd either need a truly vast singular yard like Beacon Park used to be, or need to keep the functions well-separated in separate yard spaces. CP Yard to the south could easily be fully reactivated to absorb all of North's functions and then some, but you'll be using the MA 135 grade crossings more (not a bothersome amount more, but a bit more than today) with that site.
Yeah, in my mind, the station absolutely has to be east of Concord/Hollis and needs to accommodate an elevated extension to the Framingham secondary for future :30 service (brown is elevated structure and blue is potential development parcels). I think it would make most sense for this station to have two island platforms to further avoid conflicts between Worcester + Northborough Regional Rail every 15 , then Amtrak expresses (Lake shore+ Springfield and beyond) like bihourly one day? This will be a very busy stop one day.