I bet if you knew more about juncture, you'd regret the time you'd spent learning about it. It's an old fashioned theory which might or might not be accurate (I couldn't say), but it doesn't really make sense of the facts. It's also formulated to fit the peculiar methodological ideas of mid-20th century American structuralism, which require you to be able to deduce the phonemic description, including junctures, from observable facts. I like the description in SPE (The Sound Pattern of English) much, much better, and it has had much more influence in modern linguistics. In the SPE theory, "junctural" phenomena follow from speakers' understanding of syntactic and morphological structure.
We can take your example "an aim" versus "a name" to illustrate a non-junctural description. The phonetic difference between these two is the lenis "n" in "an aim" versus the fortis "n" in "a name". Lenis means weakly articulated, and fortis means strongly articulated. Lenis articulation is typical of syllable offset consonants in English and in many languages, and in some American dialects leads to flapping of the "n" in "an aim", [əɾ̃ẽj̃m]. On the other hand, "n" in "a name" fails to flap, because it's fortis, being at the beginning of a syllable.
And why is one "n" at the end of a syllable? Well, it's obvious, isn't it? It's at the end of a word. But back in the era of juncture and all that, this would have been regarded as a methodologically unsound statement, because when you're observing pronunciation as an analyst, you don't know until after you've figured out the phonemics where the words begin and end! Thus, the need for junctures, which were thought of as being phonemes, or at least like them.
So nice to have all that behind us.