Eakin and Akin appear to be derivative names, so a match is not unexpected. If they are predicted to be R1b, it would be a good idea to upgrade. It is such a common haplogroup that you could easily have a random match.
I would not bother with the Hill match. Since you have traced back to an immigrant ancestor, any common ancestor would probably be in Ireland and it would be almost impossible to trace.In R1b, people get a lot of random matches with other surnames.
The only exception would be a highly unlikely non-paternal event in his lineage. You may want to test a person who descends from another son of the immigrant to eliminate that possibility. However, the match with the Akin, if it holds up, pretty much preludes that. A non-paternal event in genetic genealogy is when a son or unmarried daughter uses a surname other than that of his/her biological father.
You should also search www.smgf.org. It uses the same lab as Ancestry, so you do not have to change the results.
You should also enter them at www.ysearch.org. However, you have to subtract 1 from Y-GATA-H4. I would not enter DYS441 or DYS442 and Y-GATA-A10, if he upgrades, because they have become corrupted. If you get a close match, you would need to contact him to find out whether he put in the original FTDNA results, since Y-Search uses a different standard than FTDNA.
I would deal only with same- or derived-surname matches unless there is good reason to think there was a non-paternal event. About all you can really do is search with the other surname on the Eakin/Akin boards and vice versa to see if anybody has run across a connection.