Is EA reassessing its Bayesian Priors?

Given the pretty spectacular disasters that very prominent and presumably important members of the EA movement have wrought (SBF, Holden Karnofsky, Helen Toner), is it time for the EA community to rethink its basic assumptions and logic?

Two correlating questions: 

1. Could it be, perhaps, that EA's foundations are fundamentally flawed, resulting in unpredictable negative externalities as it grows?
2. If EA is NOT reassessing its base assumptions, why is that not in itself an extremely ineffective strategy for being altruistic?