I think Worcester, Fitchburg, Reading and Fairmount are the best lines for DMU-ing/EMU-ing. I'd cut the Worcester one off to Riverside rather than run all the way to Framingham. Fitchburg could use it to 128 because Belmont is loathe to allow rapid transit through (Red or Green). Fairmount is pretty much set up for it, and can't be converting to rapid-transit. Haverhill will get booted to the Lowell/Wildcat alignment eventually, especially when the N-S Link comes around, meaning Reading may as well get "Fairmounted" as a way stop-gap, and probably eventual conversion to Orange.
The Eastern Route doesn't really need it, and the Blue Line would be higher priority for improving that corridor. I don't think that the Old Colony can support DMU headways plus Commuter Trains until the the corridor is double-tracked in Dorchester. Needham can't support the headways either because the NEC takes priority over anything routed that way. Needham Line needs a Green/Orange pairing each taking half of the route. The Franklin and Stoughton lines could probably support it if they were routed over the Fairmount Line rather than the NEC.
Yeah. Your three other applications with the capacity to handle it are:
1) Worcester Line to Riverside -- SS, BBY, Yawkey,
Allston/New Balance,
Newton Corner, Newtonville, West Newton, Auburndale,
Riverside.
2) Western Route to Reading -- NS, Malden Ctr., Wyoming Hill, Melrose/Cedar Park, Melrose Highlands, Greenwood, Wakefield,
Quannapowitt/Route 128, Reading.
3) Fitchburg Line to 128 -- NS, Porter Sq., Belmont Ctr., Waverley,
Beaver Brook, Clematis Brook, Waltham Ctr., Brandeis/Roberts,
Weston/128.
(infill stops in italics)
--------------------
1) All Worcester needs to run dense overlapping service is more numerous crossovers inside 128. Right now there are none whatsoever between the west end of Beacon Park and Wellesley Farms, which is why the Newton schedule has to be so limited. If you did 1 set of crossovers per every 2 stations the thru trains would have ample opportunity to pass. And maybe keep one of the ex-Beacon Park lead tracks as a thru 3rd track around Allston/New Balance, since that would act as your "once every two" passing opportunity on the Yawkey-Allston-Newton Corner stretch.
I wouldn't run this all the way to Framingham because the gates at the grade crossings will be down way too often, and those crossings may be physically impossible to eliminate. It's also a pretty long trip
on the schedule; you ideally want your "Fairmountings" to hang close to 30 minutes. All-stops (except Riverside) to Framingham probably still pushes 45 mins. even if the speed limit on the line gets raised because of 13 stops en route. Better to cover with overlapping service patterns: Riverside locals, Worcester directs that skip everything inside 128, Framingham locals that maybe skip a couple low-demand inside-128 stops, and Worcester super-expresses that run nonstop to Framingham. Mix and match on whatever service hours to fit the demand curve.
2) Nothing needs to be done to Reading except to permanently boot the Haverhill Line back to its former Lowell Line routing and reinstate the Salem St. stop on the Wildcat Branch to replace North Wilmington (Salem St.'s a better location than N. Wil. anyway). Reading today is about 30 mins. on the button. Figure a faster-accelerating vehicle can make the same time with extra Route 128 stop infilled. Possible consolidation of Wyoming Hill and Melrose/Cedar Park if you can sell town of Melrose on the idea (I doubt it).
3) Only major thing the Fitchburg Line needs is that 128 stop in the sand pits/industrial park at Exit 26 under the Route 20 rotary. Which would knock out Kendal Green, Hastings, and Silver Hill in one fell swoop. The rest is pretty much ready-made, and there wouldn't need to be much crossover work here because the thru Fitchburg schedule isn't too dense for overlap. Beaver Brook (at Beaver St. grade crossing on the 554 bus), and Clematis Brook (under the Route 20 overpass on the 70/70A...
disused platform still there) were regular stops until 1978, dropped for schedule savings when the line was re-extended from South Acton to Gardner. Those can be reinstated. Alewife probably isn't necessary because of the Red transfer at Porter, would be somewhat of a walk, and would require a grade crossing of the T's maintenance yard tracks to exit the platform. Note that when they finish the ongoing signal work next year the speeds are going to pick up noticeably and it'll hit Brandeis a good 3-5 mins faster than today.
Eastern Route is no-go because there is not enough capacity through the Somerville track merge to mix clock-facing Eastern and Western Route DMU's. You can do it with Reading because those trains can easily hold for a passing Newburyport/Rockport train without undue schedule penalty. But start doing both and Newburyport/Rockport schedules have to be cut back. And that is not acceptable with the ridership.
Old Colony doesn't have the capacity through the single-tracking, and is superfluous because the Red Line already covers it with consecutive transfers at JFK, Quincy Ctr., and Braintree. Give up dreams of forking to Randolph Ctr. or Weymouth...it ain't happening without a billion dollars of construction around Savin Hill. Get better bus transfers to cover the need.
Needham is impossible because of the NEC congestion out to Forest Hills that's only going to get worse, and people will not ride a high-frequency dinky that gives up their one-seat with a forced transfer at Forest Hills. This has to graduate to the rapid transit system...there's no other way.
Lowell Line certainly has the capacity, but has such wide stop spacing to begin with that the upside of DMU vs. push-pull is much narrower. It is only about 27 minutes to Anderson as-is on track that is hardly as fast as it could/should be. GLX obviates any need for inner infills...and Woburn in the Montvale Ave. area is the only other place that really needs a stop (trade in Mishawum for that one). Plus, if you could get town of Winchester to stop screaming about it they could arguably push for outright elimination of Wedgemere if a Winchester Ctr.-West Medford-Route 16 bus ran at high enough frequency to make the GLX transfer at 16 superior to the CR schedule. I would say service frequency is adequately served by more West Medford buses looping at Route 16; booting Haverhill trains over to double-up frequencies at Winchester Ctr., Anderson, and Wilmington; adding that Woburn infill to both Lowell and Haverhill trains; and long-term addition of NHDOT Concord service expressing straight from Anderson to NS. By 2025 Anderson could be about as close to clock-facing service as a regular vanilla push-pull CR stop could ever be.
My question is...is this enough to justify buying a DMU fleet with the market for that vehicle still so immature? Electrifying Fairmount probably doesn't cost much more than a unicorn fleet when Providence and RIDOT South County CR would pool all the same electric equipment in much greater scale. Worcester isn't that big a deal either if they wanted to stage the wires buildout Riverside first, Framingham second, Worcester third. And the 2 northside applications already fit neatly within a 30-minute schedule slot with no physical plant modifications whatsoever, so what's the X-factor for using different equipment? Yes, there's a performance difference...but how big a difference would make people sit up, take notice, and take the train when they don't today? The trip time from Reading, Waltham, and points inbound already beats the concurrent buses and is well within the realm of convenience. All those riders want is real frequency and actual stops with parking at 128 where those lines currently have none. I think push-pull works just fine for the first 15 years of the Reading and Waltham rollouts.