There have been several observations about peer review elsewhere and I’d like to link to three excellent resources that look into this problem.
It is difficult, if not impossible, to ensure that the authors have best interests in mind. I have a potent reason to suspect that bulk of “self—authored” articles are actually “outsourced” (how that happens, requires a deeper investigation).
More than the peer review, I have been rallying for a simpler approach to the subject. We don’t have to subject the readers to a painful enunciation of statistics wherein the readers (and the audience) parachute themselves in midst of complex conceptualisations.
My reasoning is that the research should be applicable practically and unless we quantify qualitative metrics, it wouldn’t help much for the cause of science or medicine.