I agree that from a purely Bayesian perspective, we should almost certainly have larger 'belief updates' from people who are closer or have first-hand experienc

I agree that from a purely Bayesian perspective, we should almost certainly have larger 'belief updates' from people who are closer or have first-hand experience about a given topic (and I certainly found GP's post both insightful and interesting). What I'm worried about though is the risk of going past 'we should listen to person from group X when they talk about group X's experience', to the the much stronger claim that 'for all persons P outside of group X, P have nothing valid to say about group X'. That would be to confuse ends for means, and to discard the attempt to find universal truths and merely settle for truths that are relative to a person's group identities.