I read the linked paper and it's of a very poor quality. They fit an epidemiological model to the current data and then extrapolate it blindly to the future. (T

I read the linked paper and it's of a very poor quality. They fit an epidemiological model to the current data and then extrapolate it blindly to the future. (The fact that it worked in the case of MySpace is not as impressive as it sounds since by 2014 MySpace was long dead.)