More Questions for Discussion

What is the best possible outcome of a generative model of shape?  Is such a thing worth pursuing? How useful were Zhu' experiments and what others might be worth doing?