While telco lobbyists (or paid "think tank" commenters) have a long history of pushing the totally bogus concept of an "exaflood" of traffic
that will take down the internet, more recently they've been pushing this idea of "bandwidth hogs." That is, they say that even if there isn't really a big threat to backbone bandwidth (which they had claimed originally) there is still a problem with "bandwidth hogs" at the last mile, using up way too much bandwidth. And, for that reason, they insist that ISPs should be able to cap and meter broadband, to make sure that the "low level users" aren't subsidizing the "bandwidth hogs." There are two big problems with this claim. First... in none of the experiments with metered billing have the "low level users" received a discount. Instead, they've kept paying the same amount, and it's just that the ISPs have tried to jack up the rates on higher bandwidth users.
But, an even bigger problem may be that the very idea of "bandwidth hogs" may be a myth
(found via Slashdot
). Benoit Felten is smashing that myth, in noting that there are certainly some folks who use more bandwidth than others, but contrary to claims from ISPs, he hasn't seen any evidence that they're causing any harm or congestion on the network. So he's presented a challenge to telcos to send over data that he can analyze to prove him wrong.