Reddit reviews The Atlas of North American English: Phonetics, Phonology and Sound Change

Word-final [t] glottalization depends heavily on the dialect. Atlas of North American English is a really interesting resource if you're interesting in this sort of business. You can probably get it in your school library's reference section.

But there are a number of reasons against transcribing it as a glottal stop:

• Even in dialects that glottalize word-final [t], it's not always strictly realized as a word-final glottal stop. Lots of folks who glottalize nevertheless articulate the [t] (and simply don't release it, glottalizing instead), resulting in a downward F2 transition that resembles that of going towards [t]. So cases where this happens, a stricter phonetic transcription would be of a glottalized [t] than a glottal stop.
  
• The most salient reason is simply that not all folks glottalize. I don't, and we often need to "lowest common denominator" our phonetic transcription that some level. [t] seems to be suitable, unless we're in a context where the nature of the stop is specifically under question.
  
  To address the other main issue of contention of the phonetic transcription, the flap, I happen to disagree with it being transcribed as a [d] instead of a flap from a linguistic perspective. The conventional environment for intervocalic alveolar stop flapping is simply before an unstressed syllable. You could get away with saying that [d] does not flap if you were to indicate the second syllable as having secondary stress. Otherwise, you'd be implying that there is some other phenomenon that blocks flapping in this environment. Nevertheless, there's a very obvious argument against transcribing it as a flap in that visitors unfamiliar with the IPA would just get confused. That's fine with me.
  
  Bottom line is that phonetic transcription can never make everyone happy. All it should ever seek to do is to be strict enough for the context, and for this case, that's certainly the case. We could start insisting that the [r] be labialized, the [E] be semi-lengthened, the [d] be flapped, and the [t] be glottalized, we'll never end up hearing the end of it.
  
  Attachment: Imgur gallery. This is my pronunciation of Reddit, with stress patterns 01 and 10. In 01, we have a full voiced stop closure of around 45ms (no flap), and in 10, we have a 20ms intervocalic distance that can only be a flap. In both cases, I have an unreleased (hence lack of a stop burst, or even a stop, for that matter) and unglottalized alveolar (hence the slight tapering in F2) coda.

I actually met/talked to him recently- he says he's not planning on publishing anything directly from the surveys right now, as they were primarily designed to help him teach his Dialects of English class at Harvard. The page does say he's working on an atlas though, but those things take years and years to put together, and are usually very pricey.

There is a Phonological Atlas, if you have $900 to spare. The Dictionary of American Regional English (which was just completed, after about 50 years of work) is also quite fun, but similarly expensive.