It seems like an unusual problem. Zagat reviewers[1] were volunteers, but they remained anonymous so there was little opportunity to promote themselves. If I remember people seemed motivated by getting a guide sent to them and feeling like they were part of an in group. I don't know what would work now to avoid fake reviews and reviews promoting the reviewers.

It would be interesting to see what happened if it were limited to a subset of people who used OpenTable for minimum number of reservations per year or some other metric, and then was based on a Keynesian Beauty Contest[1], where a consensus is divined by reviewers earning prizes for correctly guessing which restaurants other reviewers like.

[1] https://books.google.com/books?id=CfS7BAAAQBAJ&pg=PA46&lpg=P...