In general the alignment is quite narrow, and finding space for an extra track in many sections would be... difficult.
My preferred solution would probably be:
Plymouth or Kingston, pick 1. (I vote for service into downtown Plymouth, or at least closer to it. A ~2mi extension to Lothrop St would be very easy.)
Kingston has been a great big TOD failure with the carpocalyptic big-box proliferation surrounding it, the intentional access disconnection from the adjacent country club residential, and Kingston Collection being a textbook dying suburban mall. Cordage Park, itself a TOD failure for years because of stalled development, is at least starting to get its act together on mixed-use density, and Downtown Plymouth is very walkable. I think the T needs to have an exit strategy in mind for the Kingston sand pit in the next 20 years, because the dev winds are blowing away from it and towards Plymouth. They can keep the Kingston Branch for the layover yard, but a new mainline station (closer to Downtown Kingston anyway) with all schedules running to Downtown Plymouth should be the goal.
Consolidate the RL at JFK/UMass to allow for 2 CR platforms. As part of this the split point would be moved to Savin Hill, freeing up room for full double track along the Dorchester segment.
I diagrammed out how to do it in
this post. Includes treating the area between Southampton Yard and Columbia Jct., and compacting Columbia Jct. YMMV on whether a 2-track Red mainline through Savin Hill with a split down on the Clayton St. approach is doable. Just beware that no traffic modeling has been done for that, so it might unduly fuck with overall RL resiliency. My MS Paint drawings preserve the Columbia split and 4-track JFK station (there's plenty of room to turn the CR platform into an island by reshaping the inner busway and freeing up the Old Colony Ave. pavement in a reconfig of Columbia rotary).
Double track between Braintree and Quincy Adams, there's room for it already. This leaves a 3 mile stretch single tracked.
There's room from East Braintree wye to Braintree Station if the Route 3 overpass were widened. Then the stub track of the freight yard can become the second main.
It's much tougher north of there. You'd have to nuke/rebuild the 1980 track underpass of the Red Line and associated incline because that most definitely is only single-track width. Probably also straighten the alignment so it's not such a speed restriction (it's laid out the way it is because of the switch to the now-defunct West Quincy Industrial Track freight spur). Quincy Adams Station itself also is going to be an incredibly tight squeeze. If it's true DT width (and I don't know if it is), it's got zero safety buffer from the foundations. It's noteworthy that TransitMatters left south-tip-of-Quincy Center to East Braintree Jct. as single track, given the difficulty involved. They said that :30 Regional Rail on all 3 branches would be safely doable with that stretch left single-track, and since they didn't try any crazy-high speed limit games on the mainline like they did on the branches that probably ends up holding up to scrutiny. Finishing Quincy Center-East Braintree Jct. may be something you defer till later when :15 Urban Rail service to Brockton is desired. Plain old :30 Regional Rail probably doesn't need it.
Close Quincy Center CR to increase average speed along the single-track section
Add a passing loop at Quincy Center by cutting into Freedom 'Park'
Bad idea. You
want Regional Rail stopping at major bus hubs. It's multimodal Viagra. QC itself can be double-tracked without undue expense by digging out the gravel embankment the current platform abuts, re-capping the Burgin Parkway kiss-and-ride, and punching regularly-spaced holes in the retaining wall from the current platform to reach the other side of the doubled-up island. Freedom Park and Wollaston were likewise left alone by TransitMatters (again...absent the speed-limit insanity on the main vs. the branches, so their traffic modeling is more likely to hold up to scrutiny), with the QC platforms acting as the timing mechanism for treating that 1.4 mi. of single-track to the north. You might need some DT here for :15 Urban Rail to Brockton, but :30-:30-:30 on the branches can probably live without.
That leaves 3 branches, Greenbush, Plymouth, and Middleborough/Lakeville. (Although with SCR maybe the line should be cut back to Bridgewater? That's a different question for another time.) The lack of recovery from the pandemic and the generally low ridership probably don't warrant <30min headways on Greenbush, but an extension to Plymouth would likely boost ridership on that branch, and MB/L ridership has actually exceed pre-pandemic ridership. We'll allow 15 minute headways on both of these lines. It seems like this is just barely possible with the previously mentioned measures and these headways.
Greenbush would do a lot better on ridership if it had the frequencies, and Regional Rail is supposed to be a region-wide transformative service tier so it shouldn't be a game of playing favorites with who gets the spoils. It's one thing if immovable infrastructure constraints force something like Needham looking from the sidelines on :30 frequencies; it's quite another if the killshot double-tracking for unlocking :30 frequencies on all the OC branches leads one to put finger on the scale for somebody's branch while punishing somebody else's branch.
If there's anywhere past-128 meriting :15 frequencies it's Brockton. Gateway city of 90K, large RTA bus system centered on the CR station, adjacent third-track siding for a layover yard, island-platformed downtown stop enabling easy turnbacks. Plymouth Line doesn't have a node of that kind of heft, nor the density of bus connections at any of its stops. So only one of the branches really checks off the utilization boxes for bringing Urban Rail :15 frequencies past Braintree. Do whatever additional Quincy double-tracking segments enable that, cull half the runs at Brockton, leave :30 service continuing to Campello, Bridgewater, Middleboro, Wareham, and Buzzards Bay.
