Well, it seems the main issue remains to define intelligence.
And it can't be that simple, since intelligence is supposed to specifically define man as a specie: hence, the definition is bound to be more or less circular, somehow.
Etymologically,
Inte-ligare means to "bind together" (or "
Inte-legere": to read between). It would be the aptitude to bind together elements that would have been neglected otherwise.
Neurologically, intelligence is simply a byproduct of brain activity. It could be theorized as the difference between input (senses and fixed memory), and the resulting output.
It's quite obvious that IQ tests don't measure or match either definitions. Their respective contexts are too specific to generalize anything about adaptation, and abstract reasoning is not necessarily the only way to "bind elements together".
By the way,
Ab-strudere (->Abstract) means "to simplify", and sometimes, "to take (something) out of context", depending on the tense you use in latin.
A simple child's drawing of a man is already a complex abstraction, far more than a realistic portrait. Does that mean the child is "more intelligent" than say... Velazquez or Vermeer?
---
As SolitaryWalker already assumed, there is absolutely no way an IQ test could predict outstanding achievements made by a man during his life, whether as an artist, an engineer or a scientist (or whatever). Because those specific achievements are considered to be the real proof of intelligence, far beyond any test the subject will pass.
For instance, if here we read that Einstein was considered to be the epitome of an intelligent man, it's because he was able to refine the theory of general relativity, and NOT because he had the opportunity to pass the slightest IQ test during his lifetime. People who consider he should have had a high IQ score only do it retrospectively, and this is a tremendous methodological mistake, besides the fact they invert causes and consequences, and thus make self-fulfilling prophecies. As a matter of fact, Albert Einstein and Isaac Newton never passed any IQ test. People who estimate their IQs are not only deluding themselves, they are dishonest epistemologically speaking.
And for the reverse scenario, look for instance at Chris Langan: this guy has achieved nothing during his existence, and his CTMU theory is full of odd metaphysical hocus-pocus and similar ignorant approximations. Thus, according to what I know and what I've read of him, I'd say I do not consider him to be very intelligent. Not at all: I pity him.
---
Now look at the test itself. Like I said, what's really puzzling is the fact you can train yourself to pass it, and thus, gets far higher scores after a while. The goal is to understand the purpose of the questions asked, to systematize them. There exists only a limited kind of combinations made by numbers or simplified graphic figures. There exists only a limited number of "IQ test games" and presentations. And once you are trained to recognize every category of combination, then you will quickly understand what the test asks.
It's not intelligence, it's a
repetitive technique.
Why do you score higher? Because you're trained exactly like Pavlov's dog, and know how to anticipate every possible answer.
Most people who pass IQ tests lose time because they don't understand the question, or because understanding it requires (at least for them) a great effort in abstractive reasoning, especially because such abstractions aren't natural nor familiar to most of us.
But people who know how to efficiently pass IQ tests (trust my experience!
), on the other hand, use only a tiny fragment of their brains: they simplify every possible context to its maximum. Yes: the idea is to think less to become more effective, to become quicker. You have to put your mind at rest, and think to nothing, just like an automaton. The goal is not to understand questions, but to reduce them to simple, repetitive parameters. This way, you can approximate the right answer even without understanding the question at all.
So, it's absurd!
A man with a high IQ is not at all trained to face complex, REAL situations of everyday ordinary life. And the part of the brain he has to use is totally irrelevant if he's an artist, or even if he has to write a serious scientific essay (complex multifactorial scenarios, not simplified ones).
IQ tests only show us how a tiny portion of what the brain could perform, especially within people who are able to temporarily "shut down" every other area. No wonder some very high IQs actually suffer from brain damage: such damage may in fact eases their task. If most of the time they become tragic failures in real life, it's not because they're too brilliant to be really understood by mere mortals, it's rather because by ordinary standards, they're morons, or mentally handicapped if you prefer. They're tragically unable to understand or postulate complex multifactorial situations where every lobe of their brains has to be simultaneously activated and works together: it's too hard for them, since they can't separate intermingled cognitive components to unique simplified abstractions.
On the contrary, if ever someday we should be able to measure or define intelligence, it will be either as a global cortex activity (neurologically; for instance the way different areas of our brain interact and thus create new circuits/segregation process/pruning of neurons, and possibly result in decipherable λ-calculus expressions), or either as tangible results we can achieve (empiric definition) in our respective domains of activity.
But either ways, g-factor or IQ tests have nothing to do with this: their scales are far too narrow, too restricted, limited and subjective to be anything but convincing. Basically, it's a meaningless waste of time, not even worth the paper it is written upon!
or should I say not "fake"
(my attempt at a "best of all worlds solution")... you may even have a bucket of ice cubes for extra fun! 
)
Even if it's against my ENFJ sis...
