I’m not convinced it will be more efficient for handling thousands of passengers [1], but from what I understand, this system (when operating as designed) will

I’m not convinced it will be more efficient for handling thousands of passengers [1], but from what I understand, this system (when operating as designed) will not stop cars inside the main tunnel, but move them to out of it into slide channels. That means a car driving from station #3 to station #33 can overtake cars at intermediate stations, passing them at full speed. One question would be whether it’s possible to write scheduling software that manages to do that well in crowded tunnels. Doing that with trains is possible, too, but would be more expensive (railroad switches are fairly high maintenance, and cause extra wear on train wheels), and scheduling train rides so that most passengers in a car stopping at a station want to go there is harder the more passengers there are in a railroad car. As to cars breaking down: they’re a lot lighter than railroad cars, so I would think that the tunnels can be designed in such a way that often, getting a broken car out of the tunnel to the nearest stop will be fairly fast, and can be done without disrupting service for long. [1] it almost certainly more efficient for handling low numbers of passengers, as moving an almost empty railway carriage takes more energy than moving a few cars. Scaling up, that advantage goes away fast, though, especially if, as is now the case, each car needs a paid driver.