That's not much of an argument when the supposedly "objective" measurements aren't measuring what's truly important anyway. I think the best you can do is probably to moderate the managers assessment with "360" feedback from everyone an emplotee works with, and similar feedback on the manager themselves.

If they aren't doing the same work, you can't make sure the same criteria are used.

Why should a senior engineer on one team be reviewed along the same criteria as a senior engineer on another team? Do you judge a kernel developer the same way you'd judge a Javascript expert? Fundamentally, it comes down to whether the person is helping their team accomplish its goals. And the best way to do that is to ask the team, rather than try to cast about for some kind of nonexistent "objective" value metric that allows you to rank every single developer in your organization.

Where I work we have very broad criteria that applies to almost everyone in the same occupation. Those are not negotiable but measurable - when we apply only tiny bits of subjective understanding