Yes, that's a very good point. What I am against is the rigid numerical scales. There are always ways to game them (even if easy-to-publish areas can be flagge

Yes, that's a very good point. What I am against is the rigid numerical scales. There are always ways to game them (even if easy-to-publish areas can be flagged, which would be politically difficult to do, there is the problem of people who manage to publish almost identical work in several journals, the problem of ghost authorship, etc.) I believe it's basically impossible to judge a researcher's work without actually reading it. That said, indeed many PIs can't be trusted to make decisions based on meritocracy, as there is the perception you mention and it's often justified. Probably the CVs should be evaluated or at least checked by someone different from the PI (for example by experts in the field from a different country, I think in Italy they do that kind of evaluation for some projects and grants). I just expressed some frustration as PI because I may be forced to hire a suboptimal candidate due to the rigidness of the scales, but what I want is not necessarily more autonomy for hiring, but simply any solution (like external expert evaluation) that will (1) give me the best candidate, and (2) not support the oligopoly of the likes of Elsevier like the Spanish system does.