The Gillis connection is neat. If you go to the project website, you can find an explanation of how the clan came to have many different male line bloodlines. One of the surnames for some of those lines is Gillis.
I have two McConnell cousins in the Clan Donald project, and their haplotype is fairly unusual with no close matches. Doug McDonald is a professor in the computer science field and has done some really neat things on the project website, including creation of some things called relationship trees. If you look at my cousins on the relationship trees, you find that the two Gillis men in the project are among their closer neighbors. This indicates that they are probably more closely related to my McConnell male line than most of the men in the project. The relationship is not a close one, and very likely predates the adoption of surnames, but it's still interesting.
Is your cousin the one with paternal grandfather John Joseph McDonald, paternal grandmother a daughter of Archibald L. and Mary Gillis, and paternal great grandparents Ronald and Evelyn McDonald? Doug said that he would need enough information about the pedigree file to check the values as reported by the SMGF database. I think that the information that I just gave would be enough to uniquely identify your cousin's file for him. So if your cousin would like to join, he could probably complete the process by e-mailing Doug with that information.
If he hasn't already extracted the test results from the database, I can help with that. In fact, I was looking at McDonald and MacDonald results at SMGF already, and could probably save Doug and the rest of us some work by starting with what I have already.
I do strongly recommend that he and others joining the project using test results from outside FTDNA put their information on YSearch and let Doug know their Clan Donald ID codes. I don't believe that his data (or yours) would make its way into the Family Tree DNA database, so others might have trouble finding him and others whose test results are from outside FTDNA. If they find you on the project tables, they can get in touch with you by clicking on your YSearch ID and going through YSearch. A lot of people with test results do check YSearch for matches, too.
I will ask Doug about whether my guess that results from outside FTDNA do not end up in the FTDNA database and therefore don't appear on people's FTDNA match pages is correct and let you know what I learn about that.
I thought I had your name from one of the earlier messages, but I couldn't find it when I checked this morning.