The problem here is that there is no bright-line distinction between "friendship corruption" and "doing high-quality collaborative research in a small field whe

The problem here is that there is no bright-line distinction between "friendship corruption" and "doing high-quality collaborative research in a small field where colleagues are cordial." I'm not saying that one can't bleed into the other -- sometimes they do. I'm saying that any approach you take that disincentivizes the bad behavior is also likely to harm good scientific collaborations as well. It's one of the downsides of peer review.