> rotations from one point on the surface to the other This is an interesting idea. You might want to keep track of the previous rotation, rather than just the

> rotations from one point on the surface to the other This is an interesting idea. You might want to keep track of the previous rotation, rather than just the previous point. You could do some kind of search over the space of unit quaternions to find ones which effectively distributed points.