I’d put it somewhat differently. I think you can have an adequate interpretation of quality for say the closest 100 colleagues if you have enough interactions. However, the skills you’ll be basing your judgement on will be different for different groups And will probably not be the same skills as their manager would be scoring them on. Communication skills will tend to dominate harder skills if the organizational distance increases. I don’t think that’s false per se, I do think that there is a risk views get unbalanced.