I was thinking the same for biomechanics (heart valves, etc). Wonder if the underlying idea could be expanded to work with other creasing topologies.

first, I'd like to see it replicated with different materials, and get a better definition of the assumptions required for it to apply

Inb4 more realistic damage in racing games.

Yeah, my first thought was games as well. If you could make this work in real time, you could do all sorts of nonsense.