One counterexample for why I think it's not always the best approach is that there are sometimes disproportionate rewards for being right when everyone else is

One counterexample for why I think it's not always the best approach is that there are sometimes disproportionate rewards for being right when everyone else is wrong. Doing so, however, requires you to take an idea seriously before it has gone through analyses so extensive that everyone else believes the same idea too.