Also having been in FAANG calibrations, even when people are at so-called "terminal levels" the performance ratings distribution rules still become a problem.

Also having been in FAANG calibrations, even when people are at so-called "terminal levels" the performance ratings distribution rules still become a problem. There are often team members that are at terminal levels, not interested in promotion, but doing well relative to their level standard (i.e. performing at or near the next levels). Those team members often won't be happy with a "meets" rating and still expect a rating and bonus multiplier above "meets." Meanwhile, had they been promoted, they'd probably be a solid "meets" and another team member that wants/deserve a promotion needs those above-meets ratings too. Ratings distributions usually aren't bound to just one single level, nor are they done with all levels as a single bucket, but within certain level groups, having a terminal level + rating distribution shaping is an issue.