Any set of rules can appear reasonable in practice when people ignore the rules when their results would be unreasonable.
That's the theory. In practice, what happens is that underpeformers who are good at puffing up their accomplishments or people who maintain good relationships with managers get ranked higher than people who are performing at an average level but are working on unexciting projects or are less capable of selling themselves (e.g. introverts or shy people).
Not at a couple of the companies I have worked for. Small teams were still expected to have average ratings that were close to the company-wide average.
The law of large numbers only works when the sample is unbiased. If the stack ranking was working as intended to any degree or the hiring procedures have any quality, no team will be composed of random people.
So, if that statement holds for your company (I doubt it), then both hiring managers are completely incompetent, and the rank process is completely ineffective.
- in small teams quotas are enforced pretty loosely. Everybody understands that in a team of five engineers all five can be good and smart
- on the other hand, everybody understands that out of hundred people in the department there will be a couple of underperformers.
- a department head may claim that all his one hundred people are brilliant, but extraordinary claims require extraordinary evidence
- usually there is a surprising consensus about who belongs to where. Yes, there are occasional clashes between the managers about whose guy is "brilliant" and whose is "exceptionally brilliant", but I've never ever seen a smart person being force ranked into a "bad" bucket