I'll use this as a prompt to get my sketch for this corner of the system finished up:
[EDIT: color changes indicate grade changes]
Comm Ave:
Comm Ave subway (shown here with both Blue and Green Line tracks) runs in a shallow cut-and-cover -1 subway until around St Mary's St, after which it does a (gradual) decline to -2 to pass under both the Gold Line and the Mass Pike.
Gold Line does a shallow cut-and-cover in a -1 subway from the portal until it crosses over the Comm Ave subway, at which point it starts to quickly drop to -2 so it can cut under the Pike. It needs about 260 feet of horizontal distance to do this, so it should be able to reach the needed depth just east of Carlton St. (Potentially sooner, depending on if the Comm Ave subway can be shifted into the northern lanes of the Avenue.) Beyond there, it runs cut-and-cover under the Pike.
By running the Gold Line parallel to the Pike under Comm Ave, the Green/Blue Line can avoid deeper tunneling.
Beacon Junction:
Just a little bit east of the BU Engineering Research Building, the Gold Line curves south while completing its descent to -3. (Particularly if the Gold Line hugs the northern side of the Mass Pike, there should be ample space here to build an industry-standard 82' curve radius.) The curve targets alignment with the short tangent stretch of tunnel currently used by the D Line in its approach to the flat junction with the C Line. Terrible sketch using Van's track map:
The Gold Line passes under the Beacon Street Subway at -3 and then quickly ascends
into the existing D tunnel. By using the existing D tunnel as a "jump off" point for digging our new tunnel, we can avoid cut-and-cover on the surface -- instead, we do cut-and-cover under the existing subway.
I'm not totally sure about the existing grades in the tunnel -- it's possible that the entire tunnel, plus some of the incline at Fenway station, would need to be dug out to make for a steeper grade sufficient to do the necessary elevation changes. But, as I've tried to indicate with the placement of the 260' scale markers, there should be enough distance between Fenway station and the Beacon Subway to make the necessary drops within the existing ROW's footprint (even if none of the extant physical track would be used).
Then at Fenway you have a flat junction for a surface branch into Longwood proper. This part is... fuzzy and has a significant impact on the intersection with Park Dr, which I haven't fully vetted. But that's out of scope for the particular question of getting between Fenway and BU Bridge.
Boston University station:
Parallel stations at/near St. Mary's Street, joined by a pedestrian tunnel underneath the street, with headhouses on Comm Ave and Mountfort Street, and probably on the northern edge of the Pike for the BU buildings there. Pedestrians cross the highway below ground level behind fare control. (Or, I mean, I guess it doesn't actually have to be behind fare control.)
The horizontal distance between platforms and headhouses is within the realm of existing downtown transfers, though the level of vertical circulation is less than ideal; the north-south pedestrian underpass would need to dip all the way down to -3 in order to clear under the Gold Line, which would require some long escalators which in turn likely increases the horizontal distance. It's the most painful for commuters going between Comm Ave street level and the Gold Line -- will require four stories' worth of escalators (3 down, 1 up), plus something like 300'-350' of walking.
~~~
That all being said, while I need to look at it more closely, I think I like
@Teban54's design better. One way or another, I think this stretch would need 2 sets of LRT tracks, plus I want to future-proof a Blue Line tunnel through here as well, and I think Comm Ave could handle 6 tracks (and definitely could if they were double-stacked). Comm Ave would be easier to dig under than the Mass Pike, and while I am less pessimistic about a Mass Pike subway station than some, obviously it would be better for everyone if it were under Comm Ave instead.
The double-back is somewhat inelegant, and the extra stop would increase travel time. But a transfer at Kenmore would be valuable, especially since it reduces the pressure to extend the Blue Line to BU in an initial build (since the Gold <> Blue transfer would be available at Kenmore, and a Purple <> Blue transfer would be made easier with an internal connection between the stations).
EDIT: Also, the Brookline Ave alignment would allow a non-revenue connection between Fenway and the Beacon St Subway to remain extant, which I do think would be valuable.
Plus, another station at Kenmore/Lansdowne would be invaluable with crowd distribution during "Code Red [Sox]" situations.