You can't make conclusions about the number >140 based on a shift of the mean. IQ is scaled so that the general population has a normal distribution. That doe

You can't make conclusions about the number >140 based on a shift of the mean. IQ is scaled so that the general population has a normal distribution. That doesn't necessarily mean some subpopulation has a normal distribution. Even if that was the case, the subpopulation distribution would still have a different standard deviation. (It would partly depend on the level of assortative mating within that population.)