Suppose you were ready to embark on a serious animation project. Would you:
Model and rig every character serparately, one at a time, or
Model and perfect the rig, et al., for one, then poke and pull at the mesh to make it into another character?
Knowing how the pros at Pixar or even the Orange team have done things is not as much interest to me as say what is a sound work flow advisable for say a typical TV commercial or a one-guy M.F.A. project.