'If prescriptivism is unnecessary in L1 acquisition, then dictionaries will only exist for L2 speakers, which isn't the case.'
Your deduction appears to me fallacious. Using dictionaries do not necessarily mean the chap is prescribing a rule. Take for example, if I throw out the word 'les betes noires', which is of course not an English word (but widely used therein!), you may well be perplexed as to the MEANING of it. That you are a native speaker of a language does not immunize you of being perplexed by the meaning. You may well guess from the context to what functional / lexical category the item belongs, but still have no idea what that means. This is when the dictionary comes into play. Take a look at Collins - Cobuild, or the Oxford Dictionary of English, both of which are something 'authoritative'. It is rather evident that these dictionaries are far more concerned with the notion of 'meaning' than they are really making out a set of rules, thus giving the reader a brief idea as to what the word means, and under what circumstances these meanings come into being.
It would be far more painstaking of L2 learners just to understand the meaning of an item in a particular context (esp. if the language is not a cognate to the languages he already know... English vis a vis Chinese for example), for which reason of course dictionary is an even more useful tool. An interesting remark to add is that grammatical rules observed in dictionaries for learners of the language seldom exist in their counterparts for native speakers. It suggests to me that 'prescriptivism'(at least at the syntactic level) plays really a negligible role in L1 acquisition, if any.
Therefore I rescue my statement by dissociating the relationship between 'prescriptivism' and 'dictionary'.
I do not deny the fact that to dig deep into 'language rules' is somehow a misfire in studying a language as much as I do not deny 'prescriptivism' is a useful way to begin learning a language. Language is in its nature dynamic, whereas rules are generalized according as the dynamism goes on. Therefore to stick to the rules appears to disregard this very nature of language. But to 'let the language take care of itself' is based on the assumption that the speaker already has sufficient command of the language, the study of which begins by 'prescriptivism', wherefore I think the 2 ways of learning can co - exist, and together play the main role accordingly.
Chomsky is not as 'egocentric' (ie. parochial, biggotted) as some may think. His theories have come to constant change, and most of which is not initiated by himself but criticisms from other linguistics schools. eg. his distinction drawn between the Grammatical Function and the Theta Role was initiated by Frantisek Danes, a functionalist of the Prague school, who had the opinion that Chomsky's PS rules somehow disregarded the lack of one - to - one relationship between the Grammatical Functions and the lexical categories, for which Chomsky, in response, devised a Theta Theory and later, the Case Theory (to explain the relationship between NPs and predicates and the inflectional morphology). Most of the terms used therein are in agreement with those of the Prage shool (but of course, save the 'Theme Rheme' relationship devised for a different purpose --- the FSP). To give him a fairer judgement, I think every academic who has some achievement tends to stick to his thinking, as he is much confident in it. Chomsky is just believing in what he believes and disapproves of what he does not. It is just like an atheist will definitely speak of burning incense for our ancestors as superstition.