I challenge you to find a mechanism that works for this and is quick enough to keep up at rush hour in the average Tokyo train station.
Additionally, people move at different speeds through the station due to factors like age, familiarity with the network, disability, bladder capacity etc.
It’s a non-trivial problem to solve, and likely only marginally improves user experience.
I suspect the reason this isn't valuable is that the feedback loop is already pretty good at high occupancy rates: when one car is at capacity, every other car is too.
Not to mention the trains stop every minute or two, and people don't, like, book that stuff ahead of time.
And they already know which train routes at which times are crowded, if they could just add more capacity they'd have already done it.
- which cars are normally less crowded
- which cars are recommended for getting off at a certain station, based on proximity of the exits at that station to each car
Google Maps (as well as other apps) also show similar information when asking for transit directions (for the former it only gives you the average on the whole train, not by car)