Singers' or speakers' own regional accents and dialects are of course exempt from any criticism.

The "folksingers' n", which I had not noticed before reading this thread, is quite a different phenomenon. Judging from the examples mentioned by Peter, I guess that it is used to avoid "unnaturally" long vowels. Whenever a vowel is followed by m, n, or r, Tim Dennehy spends the rest of the note with it, after a very short vowel. Thus, the n is a "liquidisation" of the t-sound following it. This effect is well known, normally very short, and so is the corresponding short m before p or b. (BTW: since the original Greek b is now pronounced v, the Greek use "mp" for the b sound in foreign words.)

In my personal opinion, some singer-songwriter's idea of sounding "heartfelt" rather than "artificial" (= vowels of full length etc.) is of questionable taste. A song will always be art, good or bad.

The "opera singers' m and n", equally questionable, are applied at the beginning of notes, usually high ones, to test the intonation and to reach the correct pitch by a glissando. The idea is that a consonant need not be in pitch - not all listeners agree.