Yes, experts can be wrong. But we can minimize our chances of relying upon a poor expert by supporting systems of robust peer review. A paper that survives pre-publication and post-publication review deserves more of our confidence than otherwise. Peer review is a minimal standard for separating bad studies from probably not that bad studies. Crap studies do get through. So we should not rely upon a single study or a cherry picked set of studies. We need to know all the studies relevant to some question. Nobody has time to get on top of the entire literature in some field except for experts. Thankfully some of them write review articles summarizing everything. Yet sometimes those reviews are crap. Thankfully a whole bunch of other experts will rip bad reviews up so only the good reviews survive.