What expert agreement should count for, what review misses, replication failures, publication bias, and conspiracy claims.