It's almost like you need to identify which employees are key contributors and pay them appropriately. Imagine that.
But I think the more important question is: who are these key contributors? It’s perhaps easy to identify some of them, particularly those who are louder, and perhaps it is also possible to identify those who appear important but are, in some sense, bluffing. But there are also likely those who are key but whom the system for evaluation overlooks. It’s probably better for those people if they leave as they may find an employer who can better recognise their talents (and leverage them and pay them for it). I don’t think I’d do a great job at recognising such key people even among those I work closely with but maybe I’m just bad at that sort of thing.
(I don't think it's quite as hard as you're making it out, though. Who are the people you go to first with questions about how stuff works?)
Maybe we have a mismatch on what "appropriate" pay means. If the expected cost of losing someone is less than what you're paying them, maybe you're paying them more than what's appropriate. But it does sound a bit like you're falling into the same trap as GP in assuming you have to pay everyone the same rate; I could be misreading you.
In practice, most places I've been have a few people who are "obviously" key and then a lot of ambiguity.
When asking people about the performance of their peers, you usually get equally positive feedback unless folks are fucking up.
You have a good point though. Perhaps managers shouldn't be asking for feedback on specific individuals, but ask individuals who they go to for help and mentoring.