The methodology on this is...extremely questionable.
Oh I agree but in terms of measuring homophobia in a country where ssm is legal, I can't find anything better. It's not something that is easy to quantify.
No doubt, but honestly I think anecdotal evidence is almost worth more in this case than a study which draws such erroneous conclusions from such a wide net.
As a queer person, I know I throw around f** occasionally with friends, and if I were a little more edgy I'd probably use it on social media too. So you're already going to be picking up people in the marginalized group using the slur, which will matter.
Plus using a slur isn't a perfect measure of attitudes toward another group. A bro in Tonawanda may use f** while playing CoD with his buddies or whatever. But he's probably cooler with gay people than a wealthy suburban couple in Shelby County who would never dream of being so vulgar.