So basically, evidence for declining mental ability with age + fancy unverifiable model => evidence for increasing mental ability with age.

This is the worst kind of PC science.