A Rationalist wants to make optimal decisions, and this often means going as meta as possible and striving to consider all known information: cognitive biases, general science, etc. [...] Rationality helps but data/experience is what's critical for making effective decisions. If you haven't validated your theory with experiments, it's probably wrong.

I don't think that's fair criticism. This community values doing Fermi estimates and checking whether the estimates are near the correct number. We have prediction book for calibrating our prediction abilities against the real world.