Tonight I attended a lecture at our beekeeper's assoc. meeting by Dr. Dewey Caron (University of Delaware) who is very much at the "cutting edge" of research into honeybee pests and diseases, pollination, etc.
Obviously, a 2 hour talk barely sufficed to even outline the subjects but he did spend a lot of time answering questions. With regard to inquiries about estimating mite numbers he emphasized several things such as taking samples over a lengthy period of time and utilizing various sampling methods. He particularly emphasized the importance of having an accurate idea of a given colony's population so that the mite numbers will have some relevance, especially if multiple hives are present with varying populations.
He demured frequently when pressed for specifics on numbers, citing the huge array of factors that would have to be considered but I did hear him afterwards saying something about a whole colony count (as opposed, say, to a sugar shake of several hundred bees in a jar) of 50 mites a day for a typical, large, established colony would not be unusual.