Your comments are mostly good, but I dispute the final assumption that no singleton => disaster. There has as yet been no investigation into the merits of singleton vs. an economy (or ecosystem) of independent agents.

If we were living in the 18th century, it would be reasonable to suppose that the only stable situation is one where one agent is king. But we are not.