I just spent the evening reviewing the male and female haplogroups of Benjaman’s 23 and Me matches that are predicted to be at the 3rd and 4th cousins levels.  I was hoping to find a geographical pattern that might indicate his origins. 
Unfortunately, the haplogroups of his matches do not reveal too much information because of their variety.  His male haplogroups are mostly R1b1b2a and its downstream subclades, with one I1* [European], one Q1a3a [Native American], one E1b1a8a [African America], and one G2a [Turkey and the Mediterranean].  His female haplogroups are quite varied.  These include HV0 and various subclades of H1, H5, and H7, also subclade U2, various subclades of U5, one T1, two T2b, one L3e5, and subclades of both K1 and V.  It is curious to me why his male haplogroups should be relatively similar, but his female haplogroups should consist of such a wide variety.