But that semantic issue aside, I agree with your point but would take it a step further. I chafe a bit even at the idea of "shared understanding of merit". Not that there aren't objective differences in performance, but that they vary across roles, and even within a specific type of job, different employees may have wildly different yet equally successful approaches.
I've come to believe the concept of A, B, and C players to be fundamentally flawed. The truth is that people perform differently in different situations. If you are hiring for a major brand-name corporation you obviously need some aggressive filtering to get through all the applicants, so I don't really fault the Googles and Facebooks of the world for their hiring approach. But I have always had success by shaping roles around the employee's strengths in practice, not based on some abstract ideal I'm searching for. Many of my best hires have come from unfamiliar backgrounds which I could not evaluate objectively, but whose diverse experience ended up paying huge dividends to the entire team.
I'm happier in teams where people try to promote people and work due to some shared understanding of merit, but realise that a lot of success in a group is due to its member's preferences and behaviours, and not to any objective measure of value. People should be able to concentrate on doing what they consider to be quality work, without getting depressed over not being seen as 'objectively good'.