This isn't a detailed critique, I don't have the data or the time for that. However the article talks about a different family size in Germany, & some other EU nations (2.2 individuals per household) versus (2.6 individuals per household) in the U.S. & Canada. The article implies that this changes the lines per person somewhat. Disregarding the fact that Germany is an extreme case, in the U.S. at least, family size increases at lower income levels, and lower income levels probably equate with lower internet use. I think the article's argument is very weak.
One other observation: The article's complaints about broadband connectivity to employees, due to larger business size in the U.S.- seems reasonable to me. Most complaints on slashdot and elsewhere are from consumers and small businesses. Of course I don't know how you would measure the bandwidth of Google with it's uTube. That would seem to be a third category of bandwidth, neither household consumer or business employee.