I'm OK with just getting a very few answers at first, for expediency.
I looked for and then contemplated writing an app that would optimize a car + public transportation route (exactly one of each) to park at a station relatively away from the city, and get closest to the destination in one train/metro ride (reasonable walking to the final destination is ok).
This is because driving in Paris is a nightmare, and then finding a spot to park is a treasure hunt (except if you park underground - they have a facility to elegantly get your kindney to pay for the parking).
I do not think it would be that difficult to write: the transport data is public and plentiful, openstreetmap has routing. I may get back to the idea someday.
Terretta•2mo ago
https://app.traveltime.com/search?aId=0&0-lat=40.7811007&0-l...
This web app is especially useful for finding places to live at the intersection of two commutes in dense cities. Do travel times from each office as their own shapes and look into the overlaps.
For the approach in the video, it seems likely "it can't work" thanks to the problem is noted at the very end. As he says, you'd need another dimension to correctly capture the differences in speeds between modes of transportation and connections.
This is easy to imagine: just picture an expressway loop and with the proximity of its off ramps to each other, and then the stretch necessary to displace addresses in the center of that loop. To his point, you'd have to lift that center out of the plane to get it far enough from every point on the circle surrounding it.