In the beginning of this continuing thread, I noted this was not a scientific study. However, based on the random articles I pulled from my collection of the following journals: The Genealogist, The American Genealogist, The New England Historical & Genealogical Register, The New Hampshire Genealogical Record, and The Maine Genealogist, I think it is obvious that only a small subset of the genealogical community read such journals.
The information contained in these journals are added by some researchers. As they make familial connections, they share the information, and that information finally makes it way on to the Internet, sadly and in my opinion, egregiously, unattributed. Most of the web sites I saw note the information came from a GEDCOM download. If there is any crime I despise in genealogy is people presenting information without properly thanking the writer/researcher who actually made it possible.
In any case, I've always wondered why a genealogist would assume that a family written about in say 1880, 1920, or even in 1950 hasn't been updated? Why are researchers always so quick to cite to older materials? Mostly, because they don't know how to look up the newest information (past a quick Google search, which as we can see doesn't really help).
That's one of the reasons I compiled my first book: New Englanders of the 1600s. It is the best research published between 1980-2005 on families that came to New England between 1620 and 1700. I wrote about writing such a book in 1997 before the world wide web took its dominant place in everyday life, but I can see that even after a decade of Internet, finding good, solid, and recent research is still difficult.