Miller's point being that those thousands of genes can easily be driven to fixation by evolution within a fairly short time, yet have not, and it's not clear from the GWASes yet if they're even under directional selection.
Right now, between the GCTAs and the failure to find lots of important rare variants affecting intelligence such as mutation load (eg no Swedish paternal age effect, unlike many disorders), the consensus seems to be swinging towards some sort of frequency-dependent or stabilizing selection: greater intelligence comes with some sort of fitness penalty (greater energetic consumption?) or maybe greater vulnerability to developmental disruption through poor environment and so a net disadvantage which favors poorer but more robust variants (and eventually, canalization). Given the accumulation of archaic & ancient genomes and further intelligence GWASes, we may be able to get a definitive answer to the old puzzle of why intelligence is heritable at all in the next few years.
Hence my old point about nootropics: for the effective ones, the reason the evolutionary argument fails may simply be that they require more metabolic resources which would be a fitness disadvantage but that no longer applies in the modern calorie-overload environment.
So Scott Alexander's post at http://slatestarcodex.com/2016/03/01/2016-nootropics-survey-results/ shows that the most "effective" "nootropics" have still been the ones that have existed for a long time. What do these results really mean, though? Is it possible that people are just worse at noticing the subtler effects of the other drugs, or are just much worse at disciplining themselves enough to correctly use the racetams or noopept (as in, with choline)?
How much potential is there in innovation in nootropics? What is holding this innovation back, if anything? It feels like there hasn't been any real progress over the last 15 years (other than massively increased awareness), but could targeted drug discovery (along with people willing to be super-liberal with their experimentation) finally lead to some real breakthroughs?