We ought to have a body that randomly audits journal submissions and checks the quality of reviews and editorial decisions. If they find that a reviewer's report had many complaints that were factually incorrect or that overlooked something actually already addressed in the paper, etc., then this auditing board could enact punishments on such reviewers such as publishing their names and their crimes for all to see or otherwise restricting them from reviewing again---a reviewer black list. Similarly, if editor's made boneheaded decisions that relied too much on subjective complaints or trusting bad referees instead of checking the complaints themselves, they should be flagged. If this were done in just 1% of cases, the process would instantly become a lot better for authors and the community as a whole.
The system now essentially operates in the dark. Aside from a few email among a few individuals, nobody is overseeing anything. Operations in the dark tend to invite corruption and bad behavior. Shine the light on it and cleanse this sickness now.
