I have had a quick 'play' with this idea but it does have a problem... Because most transmapped hair props are double sided the collision routines for the cloth get somewhat confused (as far as i'm aware collisions are only calculated using the back face of the objects) and also some of the vertices on the hair prop can be inside the figures head - further straining the calculations. As I said, I had a quick play and with a little carefull use of the constrained group reasonable results can be had but it takes a Looonnnnngggg time to calculate the simulation (because of the above mentioned reasons). Best bet is to give it a try. No doubt someone will start creating single sided transmapped hair props before too long for this very purpose. hmmmmm..... thinking about that.... time for a little 3ds work me thinks!