The more I think about this, the less I understand what you were trying to say.
The first problem is of course that a(n) is an English word. But more than that, every sentence of English must apply the rules of English to every part of itself. If instead of "scare" we used the Chinese loanword xia, we would say things like "you xiaed me". We would not say "you xia me" with the past tense present but unmarked, nor would we say "you xiale me".
But it is undefined. English doesn't have strong dialects like Chinese does. Southerners are more likely not to pronounce the "w" in wu, so they might say "an wugui" (since they are more likely to say oogway like in Kungfu Panda) while a northerner will say "a wugui", since they are more familiar with saying the w sound. This is only if they are very familiar with English also, otherwise they will completely ignore the distinction between a and an.
So...dialects make the question hard to answer.
The second part is a more interesting claim. But broadly speaking, for any two Chinese dialects you could probably identify two languages, one being a variety of English, that stand at a similar distance from each other. The second language might or might not be called "English"; it might be called Scots, or Dutch.
Whether to pronounce a(n) with the /n/ realized is not undefined, but it is specific to the speaker. This is a case where variation within languages named "English" is more than enough to show disparities, as witness the confused Americans who copy the phrase "an historic occasion" from British usage.
I took the position, in my original comment, that I was responding to someone who didn't speak Chinese (and therefore didn't know how to pronounce "wumao") but did speak English (and therefore did know how to pronounce "an"). I didn't think I was speaking to someone who knew the pronunciation of "wumao" but wasn't sure how to use an article with it.
I don't think regional variation in pronunciation in China can explain an English speaker pronouncing wumao or Wuhan with an initial w-. The citation form of the syllable has no initial consonant, and all sources are explicit about this. Chinese broadcast media, in running speech, still doesn't produce anything an English speaker would recognize as an initial w-. (I'm having to rely, for this claim, on the television programs I found on youtube, linked elsewhere.) There just isn't a source that would provide an English speaker outside China with an example of using w-. So I assume what's happening is that people pronounce w- because they see the spelling.
(Also, I don't see many English speakers pronouncing it Taivan instead of Taiwan.)
I find it very plausible that regional variation exists, though, because I know the Shanghainese (吴 dialect) pronunciation of 勿 is -- using a pinyin-like spelling -- "ve" with an initial /v/-like sound.