Note to Mr. Sargent: Do you mind if I borrow your sanity-check, devil's advocate role for just a moment? :-)
I had some time to look at this today, and I can't figure out what the deal is over congratulations for DNA investigative work. We always appreciate the work of DNA project administrators and co-admins, and the Hamilton Project certainly has a capable administrator and an impressive number of participants.
Checking a hypothesis that this DNA information was not a new change to James Hamilton's profile, I looked at the logs. Yep. That DNA info was added 4 August 2014. Which is in keeping with the last update of this informational page, http://www.personal.psu.edu/users/g/a/gah4/HamDNA/Results.html, by the Hamilton DNA Project administrator, Gordon Hamilton, dated June 2014.
Doesn't take anything away from the work done, but I believe most are assuming this is the result of recent toil-and-tears, deep-dive DNA investigatory work. I'm confident this yDNA lineage distinction dates from well before those 2014 updates...probably from the very first days of the project. I believe this primarily because of the clear competency of the project administrator.
Reading the introductory material at the project, it's evident Gordon knows what he's doing. From the time the very first panels of yDNA results were logged from these two different lineages, it would have taken Gordon all of about 90 seconds of a visual check to know, from the STR results alone, that these haplotypes are unrelated in any timeframe other than anthropological. He would never have grouped these results together.
Ken, to your point about level of STR matching: the average, or modal, values for these two groups differ by a genetic distance of three in just the first 12 markers; eight in the first 25 markers; and a GD of a whopping 18 in 37 markers. So I believe I can come close to guaranteeing that the yDNA mismatch was never in question. Heck, even without actual SNP testing, projections from the STRs alone would have shown that these two sets of men aren't even in the same subclade: one is I1a2 and one I1a3.
I'll echo my agreement that Gordon Hamilton knows his stuff and is doing a great job with the Hamilton DNA Project. But this development with John Hamilton's scion is at least three years old, probably much more, and it came about by genealogical and paper-trail research; not by continual study of, or new developments in, the yDNA. From the very start of the project--which I imagine is over a decade old--Gordon would never have associated the test results from these two, clearly different, paternal lines.