Harper DNA Surname Project - Results

At present we have results for 8 Harper/Harpur men from Ireland who are clearly related, all having Haplogroup R1a1a and DYS449=35 (a rare value). One of them kit 8711 has all the modal values for the group. The following relationships are all calculated from this individual. He is descended from Robert b. 1797. A third cousin also descended from Robert matches him 34/37. John b. 1805, a brother of Robert, has two putative descendents (paper trail defective), one a probable fourth cousin matches 34/37 and a fourth cousin, once removed matches 36/37. A third brother Archibald b. 1812 has two descendents tested, one a third cousin, once removed matching 36/37 and a fourth cousin also matching 36/37. Unfortunately, the mismatches are with one exception all on different markers and thus random. It does not help to determine relationships.
We were convinced that another Harpur family descended from Rev. Samuel Harpur of Drumlarney, Co. Cavan was related, but testing of two descendents who matched each other 25/25, but not 8711, told us otherwise as they were R1b. A member of that family is convinced that research, particularly in Ysearch, shows that they originated in the English Midlands. We also have results for other individuals (Haplogroups R1b, G and I) who may be related to others withe same Haplogroup, but not to others in the database. Recent additions show five individuals (22178, 21776,45457, 47817, and N31909), who are Haplogroup G2 and match each other, suggesting a likely, but previously unknown, relationship, which they are trying to verify. There are four (8011, 21762, 30820 and 36201) Haplogroup I who match each other, 3 at 25/25 and one at 24/25. They are now seeking a paper trail to determine the relationship.
Trying to group the many with Haplogroup R1b has been more difficult as this is much more common, and to date no obvious relationships except for a pair of cousins suggests it self.
Previously the maximum number of markers was 37 through FTDNA, but now it is possible to select 66. I am testing on my own family what benefit, if any, an upgrade provides. Re the Peterborough Ontario Harpers, I have found 2 individuals willing to be tested. One is clearly not related to anyone in the data base, and results are awaited for the second. I also found a Harper in Harpursville, NY supposedly a descendant of Andrew Harper of Ballybay, Co. Monaghan. His wife was keen to proceed, but he declined to be tested.