Comparison vs Description

Should we get away from the 'comparing new Englishes to old-school Englishes' way of looking at varieties of English, or embrace it as a useful tool? Thinking about whether a study that starts with "native speaker" reactions is the best way to go -- it may be preferable not to use categories of 'features' generated by NS reactions....or, it may be useful if the goal is a comparison. Is the goal comparison or description of the cultural/sociolinguistic place of English in (a) society on its own terms? That would be the question. I can see value in doing either one