Even if Nigerians were super selected, IQ is only partially heritable and there's a concept called "regression to the mean." Here's an explanation from someone on your side of this argument:
Meaning that even if you took the smartest Nigerians out of Nigeria, their children would regress towards the average IQ of the population, which in this case is 71. Two Nigerians with 135 IQs would have children with IQs closer to 105, only barely being about to outperform white people.
Either hereditarians (people on your side of the argument, Cochran, Jensen, etc) are wrong in their hypothesis about regression to the mean, or the average Nigerian's IQ is a lot higher than 71, or even 85. I'm not opposed to small differences in population IQs. A few points here or there. I'm opposed to the static full standard deviation, a ridiculous concept. Preposterous idea.