FlowMoP still relies on observed neighbor motion as conditioning, and its ability to model unobserved agents or sudden interactions remains to be further validated.