The metric for review quality here seems to be: another AI likes it and three journal editors like it. Should be kept in mind...