Gotta go with Melkor on this one. A valid IQ test assumes that the concept of an IQ is valid.
The comparison to the SATs is important though, because like many things in applied science, validity is sometimes outweighed by usefulness. Schools need the SAT because there isn't any other way of distinguishing what schools are producing the most college ready students.
As it happens, those who enjoy academia are also often those who are good at tests. To the extent of perpetuating the academic system, that's a useful set up, but it does not in any way reflect on the actual abilities of the test takers. Although there may be an exception. Students who may not be brilliant, but have excellent study habits also tend to do well on the SAT, and at college for the same reason.
IQ tests were designed to separate the good test takers from the good studiers, giving the higher score to the former. It legitimizes academic preferences, but it does not indicate real intelligence.
The fundamental argument between those two groups is that one thinks the other is basically memorizing material rather than understanding it. What they often don't recognize is that they're doing the same thing, except they're memorizing systems instead of information. You see three squares with a changing pattern and an empty fourth square. Your job is to figure out what goes in the fourth square. The first and most reliable thing you do is compare it to previous systems you've seen before. That gives us a tremendous boost over the studiers because they think in terms of information first. Then you take a test that requires you to know the date of the Boston Tea Party and complain that it requires a useless degree of information management.
Welcome to the Intuitive vs. Sensor debate.