While stack ranking sounds terrible on paper the actual implementation might soften the blow. Managers probably bend the rules a bit when it comes to the actual distribution in their groups.

For example consider two managers that both report to a VP. Each manager has 10 direct reports. If the managers and VP are all friends then the actual ranking distributions could vary a bit: one manager really has a strong team and the VP knows this. The VP also knows the other manager has a weaker team. As long as the VP's stack ranks are approximately correct in aggregate the individual managers probably have some wiggle room.

The point is politics in large organizations usually trumps strict rules. This can work both for and against someone hence some people don't view stack ranking in such a harsh light.