Chomsky was an early visionary, and his intuition was very good. I would compare him more to Freud than to Ptolemaic astronomers. He started from a single notion, and built a theory around it that was justified by anecdote.
But many of the tools and constructs that fell out of this theory are important. Consider the work with grammar. CTF is equivalent to PDA, but much easier to use conceptually.
Well, you can't deny that he contributed some ideas that PLT people make use of, but the question here is whether TGG is any good as a _scientific theory of language_ and I think the answer is no. The reason I compared him to Ptolemaic theorists is because he came up with some interesting mathematical ideas, but he doesn't care about systematically collecting data and attempting to falsify the theory. See: http://norvig.com/chomsky.html
But many of the tools and constructs that fell out of this theory are important. Consider the work with grammar. CTF is equivalent to PDA, but much easier to use conceptually.