I've read it a few times now. I'm going to do my best to summarize Taleb's arguments and respond to each in turn.
I: cognitive ability is a multi-dimensional thing. IQ erroneously tries to collapse all of those dimensions into a single dimension.
What are these dimensions? And how are they measured? People who make this assertion hardly ever go through the work of actually making a rigorous, falsifiable claim as to what these dimensions are. There have been theories along these lines (e.g. "multiple intelligences") and they are much shakier ground scientifically than IQ.
This claim is also just plain false, because if you look at the literature, you'll see that intelligence research does recognize the existence of many different types of cognitive ability. It just so happens that performance on tests of such abilities all strongly correlate with IQ! The correlation isn't perfect, which is why you see these different cognitive abilities acknowledged as distinct from each other, but they are still only aspects of some common factor (g).
Now, there are certain mental functions whose performance does not correlate (very much) with IQ. Things like creativity or reaction times. That's not really a problem for intelligence research since those things aren't a form of intelligence.
II: Real world success comes more from wisdom and conscientiousness than intelligence
I would agree about wisdom, although there's no real way to measure that. The way I think of it is: wisdom is where you are going, and intelligence is how fast you will get there. If you're going the wrong way, then speed will only hasten your destruction. That doesn't mean speed is irrelevant.
As for conscientiousness, we can measure how much that impacts life outcomes. Indeed it does have some predictive value, but not as much as IQ!
Also, if you're going to take some arbitrary collection of traits and say "this bundle of traits is more predictive than IQ" then you're engaging in a logical fallacy. It doesn't matter what bundle you choose, what matters is whether IQ is worth adding to that bundle.
III: IQ is only predictive on the lower ends of performance (because of it's use as a means for detecting mental defficiencies) but has no predictive value at higher levels of performance.
I think there's some stuff going on in the statistics that's going to require a lot more digging on my part, but I think the argument I made previously that having a greater intelligence merely gives you more options still holds.
IV: People who promote IQ research are racists.
Ad hominem. And bullcrap to boot. Whatever you think about another person's motives, it doesn't make their argument wrong. What really angers me about this argument is that in the same frigging paragraph he complains about people making ad hominem arguments against him. The lack of self-awareness is truly breath-taking.
V: IQ being "the best measure in psychology" doesn't matter if it's a bad measure
This argument is directed at people who want to cherry pick the parts of the psychology that fit their political preferences. It has no relevance in an argument with someone who's willing to toss out the entire enterprise.
Nonetheless, the limitations on the predictiveness of IQ should be considered. For instance, the fact that IQ correlates with performance in complex jobs by 50% means that it only explains about 25% of the variance in performance (I don't know where Taleb gets the 13% figure, perhaps from some math I'm not familiar with). All that means is that IQ shouldn't be the only tool you use to evaluate a candidate for such a job. It doesn't mean that there's no value in using it.
VI: IQ only measures performance in boring clerical jobs where the work resembles the test.
It does measure performance (to some extent) in software engineering. I presume it also does in medical professions as well. Are these "boring clerical jobs"?
VII: IQ isn't technically a measure
Um, what? It's a measurement of how well one performs a test relative to the average of some reference population.
VIII: Using national IQ averages as justification for excluding immigration from certain countries is irrational because it ignores the variance of IQ within populations.
Totally agree. For the record, I'm not on-board with such a simplistic policy. However, I do think nations have a responsibility to consider the effects their immigration policy will have on the native populace, especially with regard to issues like the strain placed on (or support given to) the welfare system. If IQ does predict a individual immigrant's likelihood of ending up on welfare, then it should be considered.