Reminds me of the Banach-Tarski paradox, where non-smoothness breaks conservation.
Isn't B-T just a consequence of accepting the axiom of choice and performing some pathological decompositions of a ball? What do you mean by:
>non-smoothness breaks conservation