> Even if peer review does not solve the problem of bad research completely, it certainly mitigates it better than not doing anything.

This isn't self-evident. In fact, it's conceivable that peer review could actually lower the quality of published research by delaying paradigm shifting work from seeing the light of day.