I have no doubt that you are familiar with his approach - else you wouldn't have been intrigued by my opinion!! Basically, what I think is that he ends up admitting that there is a "langue" in the Saussurean sense, simply because he thinks that readjustment rules tend to produce the same operations in all individuals who speak a language, and that such operations are only those that we can verify in speech considered "standard" nowadays, and that going any further is "breaking" language - he also thinks that universal grammar is tied to the brain processes we can verify in languages. While this is not necessarily conflictive or incoherent, I think languages are TOO socially regulated - the points he states about this (pidgins, translation, children learning) are more about how human societies evolve than about language itself - and if you accept things like "all languages have a pronominal system", then you have to accept language stems from a limited notion of society (abstraction of "subjects" and interaction with them), serves to expand it, and THEN society stops language in order to assert itself and limit individuals. I have the feeling that Chomsky thinks that language ultimately serves a purpose in reference to "things", and "universal things", that there's "something" behind language and its processes, which is very difficult to sustain, and all this turns transformational linguistics into not emphasising the role of individuals in all social processes and language (which would be a coherent view for an anarchist like Chomsky), but into just a cover for making "scientific linguistics", which is ultimately dissappointing.