Re: Lower Broadway and Sweetser Circle
After studying the TransitMatters video some more and thinking some more, I'm not so convinced that a lower Broadway bus lane is necessarily necessary, although it's clear that something should be done about the multiple minutes through Sweetser Circle (minute 11 through 14 or so of the video).
There are five signalized intersections/crossings of Lower Broadway: Bowdoin, Beacham, the pedestrian crossing, Dexter, and the power plant / water and sewer commission.
An oversimplified model of the major traffic flows could consider simply the Beacham light, which sends two lanes of traffic flowing along Broadway during one signal phase, and a single lane of Beacham to the two lanes of Broadway during the other signal phase, and the Dexter light which involves similar phases and lane counts.
During the time a traffic signal feeds a single lane of traffic into two lanes, those two downstream lanes seem to tend to be uncongested and flow relatively freely.
I think if these five signals are not already coordinated that upgrading them to be coordinated would be valuable, and given that some are in Everett and others are in Boston, letting MassDOT run all of them would make sense (it's nominally a state highway, so is MassDOT already running them)? I suspect in the morning peak it's desirable for the power plant light to be set up to give Broadway traffic a red light only while traffic from Dexter is arriving, and likewise at the mid block pedestrian crossing to only give Broadway traffic a red light while traffic from Beacham is arriving, and the amount of green time that traffic from Beacham turning left onto Broadway gets should maybe be a bit less than double the green time that traffic turning left from Dexter to Broadway gets. Also, the Dexter light should probably give Broadway traffic a yellow light just as the traffic from Beacham starts arriving (or would arrive if it wasn't being stopped by the mid-block crossing).
It does seem that this coordinated signal design doesn't really interact well with transit signal priority: if optimizing for maximizing total vehicles per hour across the bridge, signal timing downstream of Beacham probably should provide yellow lights at the start of the arrival of the one lane worth of Beacham traffic taking up two whole lanes (except at the power plant, where the traffic from Dexter starting to arrive should be when the yellow light can appear if there's a vehicle waiting to turn left or pedestrian waiting to cross). It might be possible for Beacham itself to delay switching from the Broadway mainlane flow to Beacham->Broadway phase in response to a bus approaching, but if we end up not having a bus lane through that narrow section at the north end of lower Broadway, the total Sweetser Circle to Beacham travel time may be long enough that having the bus pause for 15-30 seconds somewhere between Sweetser Circle and Beacham might be inevitable.
I'm also wondering how peak hour traffic counts across the highway 99 Mystic River bridge compare southbound in the morning vs northbound in the afternoon, given that Rutherford Ave sort of looks like the continuation of that road and has asymmetric traffic flows (though of course interactions with I-93 are probably significant in that asymmetry as well).