This is a piece I don't know how to explain coherently - cities function _as poorly_ for passenger vehicles as they do for bicyclists.
Tons of queueing time, low reliability, noise, hassle, expense, lack of perception of progress.
I think a good city engineer would:
1. start gathering benchmarks for travel time by mode across the city, at various times of day. 2. Start counting intersection vehicle through-put 3. Start counting intersection "vehicle dwell time" (an intersection that has 15 cars stopped waiting for a light to change is wasting an incredible amount of time)
And then with this data, start modifying how the intersections are laid out, run experiments. There's 50 things someone could experiment with, some of which would be worth retaining and expanding.
American traffic 'planning' is so boring and so openly ineffective. It's a traumatizing institution to be a part of, if you have a sense of awareness and responsibility and agency.