I agree on this, Luce, that understanding Classical literature is deepened by understanding (or trying to understand) pronunciation. But then understanding the literature of the next fifteen hundred years requires an understanding of alternate accent(s). If you spoke Latin c.-100 to +100 to Latin speakers after AD 500, say, "you'd sound downright awkward", too. It's all relative.
Also, I, an Irishman, speak with an unusual accent (as I've been told many times by people here) and yet I am an Irishman. And did Shakespeare (coming from near Birmingham) speak with an Ulster, Munster, Connaught or Leinster accent? Because they are quite different. The fact that some have written that Shakespeare spoke in a way that, in some respects, sounds closer to certain contemporary Irish dialects than some contemporary English dialects doesn't mean Shakespeare spoke in "an accent nearly identical" to an Irishman. His contemporaries might have noticed that and said something about it, surely. (That's not totally facetious, by the way, because it draws attention to the question "did Tudor English people sound like Irish people" and to the answer "surely not".) Have a listen to the recording (not that it's definitive but it's a serious experiment) of "The real sound of Shakespeare?"
http://news.bbc.co.uk/1/hi/magazine/4694993.stm
and say that sounds like any contemporary Irishman. Some sounds have an echo but generally you have to say, "not really". There are some Irish accents speaking English and I guarantee you would not be able to understand a word of them. My wife is from the West of Ireland and has great difficulty understanding the way some of my friends in Belfast talk and Belfast is only 190 miles from Galway. My point is that, if you imagine a notional standard (such as Shakespearean English), you soon discover that not only does it change in time, but at any one time there are huge divergences from the norm, and there's no good complaining that people have got it wrong. And the beauty of language and literature is that, to their users and audiences, they don't "only make sense if you understand their origins". These are immersive experiences and language shift is usually imperceptibly slow, so most people (apart from linguists) don't need to concern themselves with origins, and where and how the changes occur.
It is right, generally, to recommend a classical model of pronunciation to someone who is learning Latin and aspiring to an understanding of classical literature (although, personally, the Renaissance is my thing). I love classical Roman models, too, precisely because they can be helpful in understanding classical Romans!! but the academic uncertainties over details in classical Roman accent are significant, too!! I just think you can't tell centuries of people and big bunches of those around today that their Latin accent is "wrong" when they don't sound like a classical Roman (unless they are trying to, of course). Because accent, in conversation at least, isn't set by rule but by conventional practice, and that varies from group to group -- and it even varies within the Reformed classical notional camp. Millions have been beaten around the head in schools for centuries to encourage them to lose their "common" or "country" accents because they are hard to understand when you're not used to them. (It's not that long ago historically that schools stopped beating elision in written and spoken English out of pupils.) Accent differences have always been big. They've also been beautiful. Save the differences.
