I don’t see that anywhere in the paper?
In fact, how could they have proposed using anything to estimate FTP back in 1992?