FWIW...here's the MS Paint renders of how you'd DT just the platforms at JFK and Quincy Center, no extras. Yellow shade is the existing 6 ft. wide full-highs, mustard shade is the doubling of width to 12 ft. wide. CR track is purple, dotted purple is new track, and switch turnouts are indicated. Egresses are shown in red.
JFK
Old Colony Ave. is deleted (shown greyed-out and barricaded with black) on the slip ramp into the station busway. This assumes that Morrissey Blvd. is going to get parkway-tamed so the rotary gets greatly simplified. That's where most of the room comes from for the track turnout. Berth #2 and the idling spot on the big busway get deleted (berth #1 is OK as-is), and compensated by modding the rest of the somewhat inefficient busway layout. Plenty of space since the station "plaza" (such that it is) is expendable for something laid out a bit better. Slight mods to the ramps + ped overpass as needed, but they are likely minimal. The east end of headhouse #2 would need to be demoed/rebuilt so it can bowleg over the platform/track expansion, but most of that 80's-construction building is salvageable. I-93 overpass pegs are not shown, but there is confirmed (Street View gives an open look from Sidney St.) weaving room between pegs for doing the turnout. No mods whatsoever to the Red Line side required.
Quincy Center
Current retaining walls bookending the CR side of the station are drawn in thin black. The wall stays for the platform doubling, but has regular ped archways punched through it (say...every vestibule stopping spot on the platform) for passing to the other side. All of the expansion area is loose gravel fill, including where the sidewalk platforms over the works abutting Burgin Pkwy. South egress stairs/elevator would be modded for the platform extension. North switch would be a slower-speed turnout because it's pinned in fairly tight by the Dimmock St. overpass. South switch is not constrained. Again assumes no Red Line mods.
Slot gains would come from "running against the taillights" on the combined mainline whenever possible. For every JFK-stopping Greenbush train, have a leapfrogging M'boro or Plymouth immediately follow it in the same direction so there's an easy on-platform overtake in the same inbound or outbound direction. For every Quincy-stopping M'boro/South Coast train, have a leapfrogging Greenbush or Plymouth right behind it. Thus, you'd essentially be managing all the mainline single-track the same way as today, except "doubling up" the trains by direction to increase the supportable traffic density to the absolute max. The rest of the time, let the chips fall with opposite-direction meets to maintain strong OTP, especially when a South Coast super-long is potentially misbehaving by running late. In the end you will get much fuller all-day schedules because the absolutely painful midday gapping on the OC will no longer be the rule, as each "double-up" slot introduces +1 unidirectional trips to dole out equitably amongst the 3 Quincy/Braintree-splitting branches. South Coast FAIL, while still a shitty schedule with its own painful off-peak gaps to the endpoints and too little double-tracking south of Braintree to manage its own meets well, is still a sunk cost for its own audience but at least is fortified from tanking the on-time performance of all other branches serving the South Shore. You're going to need to backfill the mainline double-tracking in a major way before there's any Regional Railification like :30 frequencies per branch or any zero-cripple Cape appendages of course, but doing up these two stops is a very major start for not-enormous money. Practically cheap @ JFK if you can shotgun it with DCR/MassHighway for a rotary revamp. And the gettin' is good for scooping out that loose fill at Quincy Center now that the garage is completely demolished. They really want to get on that before there's any construction starts on new station air rights.