But how does parental intelligence correlate with child intelligence?
It's a really simple question, and you're not answering it. No, you're answer isn't an answer, because there is a striking difference between a correlation within a generation and a correlation across generations. The sibling correlation does not make intelligence inheritable, but you can't quote an actual parent-child correlation, so you're just assuming they mean the same thing. And you're wrong, because the sibling correlation (as measured by twin studies) only measures that component of intelligence which can be reasonably presumed to be genetic, it does NOT measure how the undefined inherited "intelligence" traits pass from generation to generation. That is the fatal flaw in your argument.
And even when you correct that fatal flaw...all the studies (the one you're mainly referring to is Bouchard and McGue, 1981) suffer from certain notable insufficiencies, chief among which is measuring the variance of an abstract measurement assigned a value that's inferred rather than deduced (namely: intelligence). And even if you managed to correct THAT, economic success, using the same methodology but much more concrete measurables (i.e. money) only very weakly correlates with intelligence, it correlates most strongly by far (>80%) with emotional maturity. So again, you're wrong...because preferentially breeding the upper economic strata in favor of welfare roles will not increase overall intelligence, because intelligence does not correlate with economic success, according to the same studies you mistakenly believe support your view.
In fact, the only publication I could find that agrees with you in any way was on David Duke's web site...which should tell you something.
And you don't even know what "Heritability estimates [for intelligence] range from 0.4 to 0.8 (on a scale from 0 to 1)" means, do you? Hint: it's different from "inheritability".