Given the test-driven nature of modern education in the US, having 1000 kids that can just read at grade level is better than having 985 kids at grade level, 5 below, and 10 above. Any resources spent on the latter would be better spent on the 5 that are below grade level.
We need to empower geniuses if we have any hope of progressing our knowledge as a species. This means a chief priority should be to find and pay special attention to geniuses regardless of social class.
That doesn't mean there is a false dichotomy where we can't also achieve baseline literacy for all; that's a complex sociological problem which might take a genius to solve.
I think this argument presupposes that denying resourcing for the 10 will definitely improve the success of the 5.
Personal side note; I’ve tended to find that exceptional people are more productive with the resources given. So I find it hard to believe the inverse would also be true.
The way schools are currently measured, the 5 are very, very important and the 10 are completely irrelevant. The only judgement is how many of the students are below the appropriate skill level. I don't believe anyone even measures how much in advance of the skill level students at or above the minimum are.