The problem is that the people who advance are not the people who are good at science, so you end up with top down direction to chase fads while promising leads

The problem is that the people who advance are not the people who are good at science, so you end up with top down direction to chase fads while promising leads get passed over. When the fads don't pan out, the PI will push their postdocs to find something to publish just to keep publication counts up, and it doesn't matter if the articles are bunk because the lab is part of a citation ring so the authors influence will only continue to increase.