This strikes me as a huge flaw in the study, and makes me question the results a great deal. It's an incredibly fluid term, and it's used differently by tumblr than it is by /r/the_donald.
So to me, all we can really read into is the vast majority being sick of the conversation period (and I'd count myself among them, by and large).
Really the only surprising thing to me in all this is how the author, ostensibly educated and engaged, could be so surprised by the results. All it would take is a 10-minute converstion someplace other than Twitter!