Once again the question seems nonsensical to me.
The crux of the quoted argument is that its conclusion has nothing to do with its premises, really nothing at all. It is as nonsensical as the argument: "The light is too bright. Ergo, my dog has fleas."
This is because the conclusion denies reliability, and "reliability" is a term of art in the world of psychological testing. In this sense, a bathroom scale is a reliable IQ testing device. Reliability does not imply valid results.
Okay, let's assume that validity is what the conclusion was supposed to be about, something like "IQ tests are not valid, they do not measure real intelligence."
If we assume that this is the intended conclusion, then the argument is valid, so that it is moot so search for "the fallacy".
Proof: An argument is valid iff it is contradictory to say that its premises are true, and the conclusion is false. The argument can be paraphrased as:
(P1): IQ tests measure not intelligence but only the ability to comprehend and manipulate symbols.
(P2): No one knows what intelligence really is.
(C): IQ tests are not valid, they do not measure real intelligence.
P1 & ¬C yields: "IQ tests measure not intelligence [...] AND IQ tests do measure real intelligence."
This is a logical contradiction.
P2 & ¬C yields: "No one knows what intelligence really is [...] AND IQ tests are valid, they do measure real intelligence."
This is a pragmatic contradiction (inasmuch as we would be required to know what intelligence really is, in order to confirm that IQ tests do measure real intelligence).
In both cases we are contradicting ourselves when asserting that the premise is true while the conclusion is false. Ergo, the argument is valid, and the question "where is the fallacy?" is moot.