Moral concerns aside (and there are many), the potential future predictions of this kind of technology are lacking in one critical element — they do not integrate the impact of Learning Disabilities on functional educational performance. While a dyslexic, for example, could have a genius-level IQ, he might not excel in a school environment unless he is given the right supports. But with the proper school supports, he could be the next Albert Einstein (who famously had multiple Learning Disabilities).

Anecdotally, Learning Disabilities clearly have an element of genetic inheritance. But it is not a one-to-one correlation, where only dyslexic parents produce dyslexic offspring, etc. This unaccounted for piece of the genetic puzzle would at least confound, if not invalidate, almost all the results.

Roughly 20% of all students have some form of Learning Disability. The numbers are just too prevalent to exclude this factor from the analysis.