Because the momenta of the particles are random. What are the odds that all of the motions with respect to the center of mass will sum to zero? In the case of momentum, we can adjust for net proper motion by a change of coordinate system, but net angular momentum can't be normalized away.
I would expect 1. It seems to me that something not moving is something not moving. I don't expect to see a salt crystal, which has a center of mass, to suddenly start spinning for no reason.