Your argument relies on the following assumptions:

1) That averages have anything useful to say about statistical outliers.
2) That educational attainment is solely related to ability, and is therefore a useful proxy for intelligence.

As for the first assumption, 1 in 10 is still a "much higher proportion" than 1 in 50, but my argument still holds.

As for the second, high tuition, helicopter parenting, and a public school system designed to serve the bright-but-not-gifted population are all well-documented socioeconomic influences on educational attainment that have nothing whatsoever to do with ability.