(Of course, I was in the middle of writing this comment when I was interrupted by a 4-hour lunch and conversations with a colleague, and I came back to like 10 more comments.)
For this comment, I'll use "Day Square", "Central Parking" and "Logan Terminals" to refer to BL Airport station, a single centralized stop near the terminals, and 5 separate stops at all terminals, respectively. "Airport" refers to a possible choice of terminating Urban Ring among one of the three.
The other thing frequent users are aware of is the convenience of the transfer, particularly with luggage.
Orange to Blue at State is not a great transfer with luggage. At least one direction requires stairs (IIRC). And the station is really spread out.
I agree with the issues of OL-BL transfer (the walkway to southbound OL at State in particular is ~580 ft long), but this is something that highly depends on the Urban Ring's design, especially at
Airport.
Day Square completely depends on where you place both the UR station and the APM station. In the worst case, as
@737900er suggested, you'll have to go up, then down, then up.
Central Parking has a 10-min walk to terminals (even with moving walkways, that's still far).
Logan Terminals is seamless for riders, but I've already discussed the political (and even route design) issues. Realistically, in a world with an APM,
Day Square is probably the most likely option until there's political will for another cross-harbor tunnel.
The OL-UR transfer is better, but still with caveats. The transfer at Assembly, if any, will likely involve a walkway that's not much better than State. Riders looking for convenience will end up in Sullivan (longer ride), and while Sullivan itself will likely have side-by-side platforms, riders will still need to go upstairs and downstairs (with the OL platform lacking bidirectional escalators).
(And if you're considering lines west of OL, now the Urban Ring transfer totally loses. Government Center and the new Charles/MGH are easy transfers, while East Somerville, McGrath and Kendall/Tech Square will likely not be.)
Those OSRs look pretty competitive to me. 11 stops directly or 9 stops with a transfer in the middle? That seems basically the same to me. If I had luggage, I'd opt to skip the transfer, probably. The only one on your list that is less competitive is Kendall, sure. But then there is the proposed UR stop in Cambridgeport, too. I'd probably ride that one seat to the airport, all 16 stops, rather than do 2 transfers. Especially if the transfer at Kendall is a 1/4 mile walk to the Red Line. With luggage.
The problem is that the majority of riders do not enter the rapid transit system at Sullivan, East Somerville or Kendall (especially the last two). They enter at places like
Wellington, Malden Center, Gilman Square, Medford/Tufts, Union Square, Central, Harvard, Davis, etc. The combined demand from all these far outweigh neighborhoods on UR itself like Cambridgeport.
Suppose Urban Ring stops at
Day Square. Sure, a Sullivan rider (e.g. from bus transfers or the East Somerville neighborhood) will compare:
- Urban Ring (one seat ride)
- Orange Line - Blue Line
But a Malden rider will compare the following, which erases the biggest advantage of Urban Ring - the OSR:
- Orange Line - Urban Ring
- Orange Line - Blue Line (likely fewer stops, shorter distance, and shorter travel time as @Riverside showed)
That's exactly why I said the T104 bus will likely prove to be more useful than Urban Ring for airport riders. And I don't think I need to even start doing the GLX and RL comparisons.
Not to forget: people may go somewhere else on the Blue Line that's not Airport! For someone going to Orient Heights or Revere Beach, now UR totally loses out due to back-tracking at Day Square.
So the question becomes: Is it worth figuring out the engineering and $$$ challenges for a better Chelsea Creek crossing, so that people boarding from Sullivan buses/neighborhoods and Cambridgeport have a more convenient ride to Airport for the couple times per year that they need it?
But as a basic premise here, do people count stops to choose subway routes? I don't think I've ever done that in any city I've been to. I generally look at a (non-geographically accurate) system diagram and see what looks most direct. Nowadays most people probably use Google maps or something similar. If there are multiple options that take roughly the same amount of time, people take whichever looks easiest, or whichever train shows up first. I'm not sure people generally count stops.
On the contrary, I've seen plenty of people in highly complicated rapid transit systems (think of those with 10-20 lines in Asia) counting stops in front of station maps. (But yeah, that's a minor point.)
But what I'm mostly missing is why you wouldn't just put a tunnel under Chelsea Creek? That doesn't even seem like a "crazy" transit pitch. That's just what you'd do if you built an Urban Ring. My guess is you'd do an immersed tube tunnel, like they did for the Pike and Silver Line across Fort Point Channel. Chelsea Creek is even a little narrower with far less existing infrastructure around. There are plenty of sites along the creek to potentially build the tubes. How is this even a questions for a "crazy" transit pitch?
I think everyone agrees that a tunnel under Chelsea Creek can certainly work in a world with unlimited budget. However, it circles back to the question I just asked above:
cost-benefit analysis.
While "cost of 170' bridge vs. tunnel" is a separate question, regardless of the method of grade-separation, it needs to have enough usage to justify it. Which is exactly my original point: I'm not saying it doesn't have merit or isn't worthwhile enough (it can absolutely turn out to be), but
the benefits of a grade-separated Urban Ring route to Day Square may be far smaller than it might seem on first glance. And as
@JeffDowntown seemingly implied, the cost of such a tunnel may also be used for a more direct Sullivan-Day Square (or Maverick) connection via Charlestown, which brings more benefits to riders doing OL-Airport, GLX-Airport etc, and is equally capable of being extended to Central Parking or Logan Terminals.
We need a deeper understanding of the corridor, rather than "oh it connects Sullivan, Everett, Chelsea and Airport, so we're good", which likely overestimates the practical benefits. In particular, we shouldn't conflate the
"radial" needs of Everett and Chelsea with the
circumferential needs of a faster connection between OL and BL. And while the Grand Junction Urban Ring can succeed at doing the former, I think that too often people assume its success will come from the latter.