Split from here. I linked, here, to a serious scholar, who is definitely mainstream and has even published a book (with some other professors) which has become the most used book to teach MSA, and who has apparently learned many dialects, as well as MSA, and holds the view that no dialect is any closer to MSA than any other dialect. Who knows, maybe that is a PC statement on her part, but unless someone (preferably familar with several dialects) has conducted serious studies and has empirical data to suggest that a particular dialect is closer than another, I would tend to believe her. Most of the time people who claim that one dialect is closer than another to MSA are generally partial to their dialect of choice and have no real proof to back that up. So, barring some empirical proof, such claims are best left to the realm of conjecture, in my opinion.