While true that reproducibility is the truly useful filter, peer review is still a valid pre-filter. Do you believe peer review suffers from both false positiv

While true that reproducibility is the truly useful filter, peer review is still a valid pre-filter. Do you believe peer review suffers from both false positives and false negatives (type I and II errors, don't remember which is which)? As a side note, we need better terminology than type I and II or false positive/negative. Type I and II is completely opaque to meaning. False positive and negative is unclear because sometimes the negative is desirable and othertimes the positive is desirable. Maybe RDO and AUO: rejected desirable option and accepted undesirable option.