John, I don't know anything yet about Peter, but from the Dinwiddie Tax records you kindly shared with me, I agree with you that there's no doubt about the fact that Harmon stayed in Dinwiddie Co, VA and probably died there sometime between 1828 tas assessment and the 1830 census. Although it's possible that he did move south, I seriously doubt it, but some of his sons may have. His (probable) son George married twice, had 3 sons and 2 daughters by his first wife before 1820, evidently none by his second wife, and is found on the 1850 Dinwiddie Census with a household consisting of himself (74), 44 yr old son Robert, dau-in-law Julia (30) and a 6 yr old grandson named after him. I don't know anything yet about George's other 2 sons. Nor can I reasonably link any of the other 1850 Dinwiddie County Abernathys with any of those from 2 decades earlier without additional information, much of which is sadly lacking. Just when I think, for instance, that John might be a son of Maston along with James and Thomas, I realize that there are too many sisters almost his own age with him in 1850 to match up with the 1830 tally codes for Maston, so have to toss out that possibility.
I am theorizing at this point that the Freeman Abernathy who is said to have married Harriet Elizabeth Holloway on 12 Nov 1847 in Greenville Co, VA was =NOT= one of George's sons: the dates don't match up nearly as well as they should.
25 yr old Freeman Abernathy was located in 1850 in the Southern District of Dinwiddie Co, VA, and his possible father, based on 1830 VA census data could have been Harold or Maston Abernathy of Dinwiddie Co; Raleigh, Hutchins, or Matthew Abernathy of Brunswick Co; or Turner Abernathy of Lunenburg Co, VA. I did not consider any of the 1830 W VA bunch, partly for geographic reasons and partly from the belief that they migrated west, not south. I also ruled out the James Abernathy in Cecil Co, MD in 1830, since I suspect he may be an offshoot of the New England line. I realize, of course, that we might not have all of the 1830 census records, so I may have missed yet another possibility.
Much of this is still speculation, even the theory that George was a son of Harmon lacks positive proof as yet and it is all subject to revision upon receipt of additional data.
Please extend my thanks to your wife for extracting that Dinwiddie Tax data on ya'll's research trip. They've been a godsend in helping unraveling this little bit of the bigger puzzle.