Does majority of linguists accept universal grammar?

The Wikipedia claim that 'the majority of linguists accept universal grammar' is highly unlikely to be true.

I would posit an alternative hypothesis:

The majority of linguists do not care about universal grammar. They may be skeptical or enthusiastic about the whole enterprise but they probably do not know enough about it to be able to adjudicate. It is simply far too irrelevant to any of their professional concerns - it only has any bearing on a linguists work if they study syntax in the GB/minimalist framework. Of those who do know enough, there may be slightly more proponents than opponents, simply because adherence to the dogma is part of the GB/minimalist (Chomskean) framework. The opponents come from language acquisition research (like Tomasello at al) or have looked at the claims and evidence out of idle curiosity and are simply incredulous at the poverty of the evidence and the porousness of the argument (like Sampson or myself). This disparity may make it seem to an external observer that universal grammar is accepted. Many of these external observers are clueless students of philosophy who - in grand philosophy of language tradition - are about 20 years behind the latest developments in linguistics.

That was perhaps a bit more personal perspective than appropriate for SE but I'm not sure in the absence of a comprehensive survey, anything more definitive can be said.

As to the comparison of the position of Universal Grammar (UG) in linguistics to other sciences, it is nothing like relativity or natural selection. It is more like String Theory. It is one contentious (rather far-fetched in many people's opinion) way of fitting some of the spurious data into a model. As far as I can tell, UG can describe a lot less of the data and unifies a lot less of what we know about language than String Theory does.

UG often goes hand in hand with evolutionary psychology which is also highly contentious and more talked about by outsiders than its true position within the sciences merits.

As to a 'paradigm' in linguistics, I think a good place to start would be Dixon's 'Basic Linguistic Theory' which describes things like those jlawler mentioned in his comments - he describes concepts that are used by most linguists regardless of their theoretical framework. There are several competing theoretical frameworks (that may count as paradigms) of which perhaps these four are the most important:

1. Functional linguistics/grammar in the broadest sense (includes corpus linguistics, discourse studies, conversation studies, linguistic typology, lexicography, etc.)
2. Generative/formal grammar (linguistics)
3. Construction/cognitive grammar (linguistics) (close to and possibly fitting under the functionalist paradigm)
4. Minimalism and UG (related and possibly part of the formal/generative paradigm)

Update: Fixed spelling of RMW Dixon.