The point of the article was that you have to demonstrate there is a problem if you want to claim there's a problem that needs solving. If the "hogs" are doing their huge downloads when most customers are asleep, and if there's no impact on the rest of the customers, then what's the problem? Simply saying that 5% of the customers download 75% (or whatever) of the packets doesn't mean that that 5% is impacting the rest of the users or the network as a whole.
ISPs need to show more data if they want to convince educated users that throttling is necessary.
Cheers,
Scott.