[joosters]

The conservative, baseline prediction is generally "things tomorrow will be the same as today"

Note that this line of thinking means that you'd never predict the rise of exceptional giants like Google & Facebook in the first place. If your line of thinking can't entertain the creation of these companies, it's unlikely to be any good at predicting their fall either.

Can you think of a possible way Facebook could lose all their traffic in that 5 year timespan?

My point is that since we can't imagine all the possible things that can or will happen in the next 5 years, the fact that we can't think of a possible way Facebook could lose all their traffic becomes meaningless and useless as a reliable source of predictions.