That's an interesting example. The stereotype that "people from the south have an accent" is actually false. EVERYONE has an accent. Every region has regional patterns of speech that a large percentage of the population adopts. But the only indicator that these speech patterns actually provide is the regional association.
Furhtermore the regional associations are hardly predictive. So called 'southern' accent is actually very close to the old british accent, and thus you cannot even use the accent to draw conclusions.
And it is completely FALSE to assert that those accents are influence by culture. They are purely audio-linguistic patternings - vocal muscle-memory (that's why singers lose their native accents and pick up whatever accent is associated with that song).
And this is precisely where stereotypes like
People from the South have an Accent
goes so horribly wrong. You USE it as a marker for culture - when it is nothing of the sort. You USE it as a means of stereotyping and drawing conclusions that you have about the imagined or projected cultural values that are associated.
And those are just that, imagined or projected. And thus the "stereotypes" associated with your projections are just incorrect.
When the variance across a subculture is equal to or greater than the variance within the general culture (and that is the case in all of thes "subcultures" you are referring to) then they are not useful means for drawing conclusions. Yet the stereotypes about these "subcultures" are used to do exactly that. And hence they are worse than useless, they are mythologies