Have any studies been done that look at decreasing turn around times at North and South stations in order to effectively increase capacity (allowing more trains to be run during rush hour)?
SSX does exactly that. Current South Station bottles up too many trains into laying over on-platform because of cross-cutting movements across switches, a result of the formerly symmetrical switch layout being chopped in half in the 60's when the station was partially demolished. And this gets worse by the year as NEC traffic increases, because as that new traffic fans out to the middle platforms they've cutting across switches that bottle up other trains on the platform. Further, every Amtrak trip has to make a pit stop to Southampton Yard for between-run chores, meaning the switches...which are already choked going towards the yards/Fairmount/etc. direction...have to get rationed for Amtrak priority. That leaves the T little ability to swap trainsets when, say, the A.M. rush-hour inbound that just arrived from Providence has 7 cars but the next reverse-commute outbound only needs 5. Capacity's wasted by the layover, and it's wasted by not being able to match train lengths with any sort of precision.
The primary thing...divorced from all the real-estate empire-building that's larded it up...that the SSX
track work does is undo the pinch at the switches. That requires, as a necessity, more platforms...but it also means you can get on/off those platforms with fewer delays. And swap trains in/out of the yard more quickly for matching cars with passenger counts, meaning more cars available to plug elsewhere with headway increases. This gets easier still if a good chunk of the southside gets electrified with EMU's, which can mash together/pull apart in the yard on their own power instead of requiring a switcher. But regardless of electric vs. diesel, this is how you get the most possible service rotation out of the car fleet in an RER universe.
North Station's issues are simpler. Its switch layout is pretty straightforward, and since Boston Engine Terminal can be reached with relative ease and few conflicting moves from about 2/3 of the terminal the trainset swaps are less of a problem here (though it still gets hairy at rush). It's the drawbridges mostly. 10 tracks (+2 inactive) have to mash into 4 for the bridges, then re-spread into 7 (+1 inactive) on the Somerville side. Too many trains have to wait with a pinch like that. But since both draws are getting expanded to 3 tracks that'll do a lot to de-gunk the interlockings. 12 platform tracks just have to go to 6, with less sorting on the other side going back up to 8. Things will get a lot more fluid after those extra tracks come online.
Ultimately for full RER service you'll need Draw 3 re-added after the wrecking ball takes the old Spaulding building down. Let's say there's 18 platform tracks at that point just from filling out the remainder of the building, and on-ground reshuffling lengthens some of the too-short platforms there. Since there's room under the N. Bank bridge for 1 more track into Tower A, addition of an identical Draw 3 means there'd only have to be a mashing of 18 platform tracks into 9 lead tracks (easy...two sides of an island merging is the least-invasive part of terminal dispatching), and then the 9 leads would carry over all the way into Somerville making crossing over on/off the 4 mainlines much more academic.
In terms of how can you speed up
changing ends...there is an upper limit there of about 10 minutes bare minimum, with allowances for cushion. And that's mostly regulatory with mandatory brake tests, trainlining tests, etc. making up the crew's between-runs. It doesn't get measurably easier with an
xMU vs. loco-hauled train, either. And changing ends is much slower on a RR than on a rapid transit line because of the time chew of the regs. But unless they're habitually wasting an off-scale amount of time changing ends it's neither here nor there because that's life under the FRA and shouting at the moon like a transpo blogger isn't going to change that. If anything, having more level boarding platforms at the ends of lines makes it a lot easier for the engineer to jog from one cab to the other...for the same reason we riders need that accessibility to limit dwells.