Yes, it's the statistically significant difference from the rates in the general population that is of interest. It suggests there is some relationship between this genetic characteristic and (identification with?) homosexuality.

Given that almost every validated study has pointed to a complex relationship between genetics and environment in the influence of sexual orientation, we wouldn't necessarily expect a majority. It seems likely that what we refer to as "homosexuality" (or bisexuality, heterosexuality, etc) is in fact multiple phenomena with multiple causes: "homosexualities", if you will. It could be that this genetic "marker" (such as it is) points to an explanation for one or more of them. Conversely, it could also be that the marker only explains the homosexuality of some of the people who have it.