I always find it amazing that people like this are able to finagle their way into meaningful positions in SV (this is also partially in response to the bloomberg article sibling comment).
In the SV investor mind he went from running a company of creative people to running a different company of creative people. What those people actually do doesn't really matter. "CEO skills" transfer between companies quite well.
The task of a CEO in a pre-product research company is pretty different, I'd imagine. The fact that he was replaced is weak Bayesian evidence in that direction.