Without commenting on the article’s argument, I nonetheless found this graph immediately suspect, because it doesn’t account for the increase in internet traffic over the same period. The use of proportion of the total as the vertical axis instead of the actual total is a interesting editorial choice.

No, of course not.