On the subject of conforming clothing, why couldn't a shrinkwrap algorithm be used to smart fit conforming clothing by scaling each body part on a frame by frame basis?

Bend an arm and the program would detect the elbow pokethrough and attempt to scale/rotate the affected clothing parts until coverage is maintained.

Also, why can't conforming clothing objects be given a render flag so that they are automatically always drawn in front of their corresponding figure object in the hierarchy?