I am not familiar with it, so I can't really say how it might fit in. Please share any further details and your proof of concept, though. Maybe it would end up being a good fit for TM in some way.
I will mention the main reason user data is submitted via .list files and why I have no plans to allow users to modify their travels interactively. The TM database is read-only for all web-facing code. This eliminates many potential security problems, and given that I'm not a database or security expert and that we can't risk any kind of security breach given than TM is housed on a server owned by my employer, using their power, network, etc.
Another thing to consider: the site update code that reads the highway data and processes .list files builds data structures that do not remain in memory after the big .sql file that populates the TM DB gets created. There are consistency issues with the DB (route concurrency detections, stats and rankings stored in the DB) and updates to traveled-format graphs that would need fundamental changes if we were to switch from the static data we have now to something that would update any highway or user data dynamically.
