The methodology and reasoning behind this study is dubious at best. http://www.google.com/trends/explore#q=myspace%2C%20facebook... If google trends is of any predictive value, Uber and Instagram should not exist (or at least be worse than myspace). The keyword 'snapchat' also shows near zero volume.
Or that browsers have gotten smart enough that they're autocompleting facebook.com instead of just letting you search for facebook.
I do love a good Google trends "research" dive but I'd hardly consider it good enough to cite for a scholarly paper. There aren't even real numbers, just relative search volumes.