Quickly jumping on to reply to some of the points here, while I have not given the rest of this proposal the attention that it deserves. Apologies in advance if I miss something because of this.
One alternative is what
@Riverside suggested as an "East Somerville alignment" for the Grand Junction line, and IMO is probably among the most satisfactory proposals in this area holistically. The idea was originally proposed
here, then refined
here, and I wrote an evaluation of it and several other possible alignments
here (which Riverside himself updated
here).
The latest rendering of East Somerville alignment goes like this:
Basically, use the freight tracks next to New Washington St (or build an El next to it) to get to East Somerville, then run alongside GLX (possibly taking over its yard leads), and turn towards Grand Junction via a short tunnel under Fitchburg St. An infill station on the D branch near Squires Bridge is typically also proposed to allow transfers.
If commuter rail infills on the Lowell Line (at East Somerville) and Eastern Route (at Sullivan) are also built, this is one of the proposals that enable the greatest number of transfers to radial lines - GLD, GLE, Lowell, Haverhill/Reading, Newburyport/Rockport - and has the highest degree of feasibility. Only Fitchburg misses out, but at least it has a built-in connection to Kendall via Porter, partially offsetting the lost circumferential transfer.
(In contrast, most other proposals completely ignore commuter rail transfers or at least have no feasible way to do them. Many of them even ignore GLX transfers, or put forward with infills that are almost certainly infeasible.)
Yea, there were a few things missed, there's a bit more nuance in my post that weren't taken into account.
1. I am not trying to connect Grand Junction with GLX, GLX from Lechmere doesn't exist on this alternate history/reality/fantasy map (notice the dead end at Lechmere), but instead all rapid transit lines converge on the
hugely absurd "mainline ROW rapid transit concept" (yes, it's an absurd concept,
but so darn fascinating). The Lechmere Viaduct on my map receives surface routes, and the hypothectical rapid transit routes instead use the NSRL built to receive rapid transit service from all mainline ROWs. This means the track layout may differ quite a bit from present day.
In this concept, the GLX spagetti junction doesn't exist,
but one may not take any new land or tunnel for a new subway, however, the current existing or historical ROWs may be widened or tunneled, or the junctions smoothed.
Essentially,that leaves only sites #5 or #6 for a relocated Brickbottom Station (site #6 can be removed since it doesn't go near the Fitchburg rapid transit line, rerouting the B & L rapid transit line would be needed anyways). I'm still not sure whether site #5 is even a proper site for a station to begin with (again, GLX and it's spagetti junction doesn't exist on this map).
Kinda annoyed by the fact that the old junction was a crossover/flyover, and the two routes never actually were physically connected. I suppose there'd be room to slot the B & L rapid transit trackage and run alongside/parallel the Fitchburg ROW rapid transit trackage eastwards to North Station.
One question whose answer would provide more clarity into this issue is: How many intra-city transfers are there in Lynn and Quincy, and what's the demand for such transfer patterns? I don't know the answer to that, but if such demand is high, it would lean more heavily towards hub-and-spoke.
Note that for Lynn in particular, an extension to Swampscott - and even beyond that to Salem - is well-justified before even considering bus transfers. According to the 2003 Program for Mass Transportation, a Lynn-Salem extension has higher ridership per mile than Wonderland-Lynn, and the only reason it hasn't been built is that the Blue Line hasn't been extended to Lynn yet.
Yea, that's something I'm not aware of how to get easy data on. It's a
bit frustrating that the T's market analysis frequencies map does not go all the way to North Beverly's bus terminal, and cuts off right after the Lynn bus terminal, so I can't easily gauge how much demand there is between Lynn and Salem/Beverly.
In the case of the Blue Line extension to Salem, Swampscott is about as far as the 441, 435, and the 455 buses can be cut back to. The Eastern ROW misses the 450 and 455 essentially completely between Swampscott and Old Salem stations, meaning that BLX to Salem would not be able to allow curtailing of the surface routes beyond Swampscott. The only route curtailing would be the 455 being cut back from Salem Depot to Old Salem, but that would cause the 451, 465, and the 450 to have longer routings out from Old Salem vs. Salem Depot. Parts of coastal Salem east of Lafayette St. are also too far (>800 meters/.5 mile) of a walk
to access Castle Hill and Pickman Park stations, and only have the 455 nearby.
Note that BNRD is inapplicable for the fantasy map, as it is essentially an alternate history/timeline/fantasy map. The bus network would be completely different if all the mainline ROWs carried rapid transit from the start, perhaps curtailing a lot more surface routes at rapid transit transfer hubs the same way as BERy did. I've essentially opted to stick to pre-BNRD bus routes on my alternate history/reality fantasy map, and optimized them a bit (or added some fantasy ideas to them).
As for the 55x series (the reworked 56 and 58 in BNRD), I also suspect the duplication intentionally provides local service between the commuter rail stations.
In the case of extending rapid transit west of Newton Corner to West Newton, the distance from Newton Corner (Centre St.) to Newtonville (Walnut St.), is 2.02 km (1.25 mi). If an ideal rapid transit stop spacing is 1km (0.625 mi), that would mean an infill station at the site of the 558 bus at Adams St. (there is a grade separated street/RR crossing there), as that site is located exactly halfway between the two stations and would fill the missing gap there (the resulting stop spacing is almost exactly 1000 meters (0.625 mi).
The gap from Newtonville (Walnut St.), to West Newton (Washington St.), is 1.83 km (1.14 mi), which is 1.8x higher than the ideal stop spacing, but there are no good spots for an infill to fill the missing gap, with a lack of a street crossing halfway between them. It's worth noting
the T's market analysis indicates very little transit demand on the south side.
During the Highland Branch conversion to LRT rapid transit in the late 1950s,
a new infill station was constructed at Fenway, This covered a 1.52 km gap (.9 mi) between Kenmore and Longwood and today provides the rapid transit connection with the 47 bus.
I just find rapid transit extensions that consume/eat away at duplicative bus routes, allows the rapid transit extension to have the most impact at redeploying bus frequencies elsewhere as needed.
