Ideally it would be something like, "we have 100gbit of combined IP transit purchased and peering with google, he.net, and herpderp" and TCP/IP would do it's thing.
Particularly when you hear of countries outside the US, 100+ mbit Internet is a very relative thing and it depends very much on where that traffic is going.
I'm all for building out last-mile distribution with fiber, but what's possible on Cable and DSL is very good and adequate for the current generation of services if the policies would match what the technologies are capable of. The elephant in the room is that good Internet will continue to eat the lunch of Cable TV and phone. Bandwidth caps are more about keeping you in line as a "good consumer" than the economics of IP transit.
Read the small print. "Traffic is unlimited. If you exceed 40 TB / month, the connection will be limited to 10 Mbps."
That works out to a bit over 100mbps commit. A fair enough deal but nothing spectacular. Doing 10gbps ports is a neat marketing trick but I can't think of too many situations where you'd balloon that large and yet still sit within 40TB/mo where a gigabit port would be a serious bottleneck.
Just keep in mind the carriers are using at most 40gbit and 100gibt optical lines and trunking them together. If you wire an entire neighborhood with gigabit last mile, or a DC with 10gig to the node, it all coalesces somewhere at the network access points. For home subscribers, I'd much rather prefer a quality product in the 10s of mbit for the next few years, the only point I was originally trying to make. The argument for laying gigabit last mile has more to do with last mile life cycle IMHO than truly giving customers a full gigabit pipe at the moment.
Then it is 0.89 per TB and that's for UK customers. French customers and customers who sign up through ovh.com get a completely different bandwidth structure. They get "unbilled" bandwidth but only get a few hundred mbps dedicated with each peer. Google it. It's due to competition from online.net in france.
10gbps makes sense if you also have 1gbps servers in their network.
Particularly when you hear of countries outside the US, 100+ mbit Internet is a very relative thing and it depends very much on where that traffic is going.
I'm all for building out last-mile distribution with fiber, but what's possible on Cable and DSL is very good and adequate for the current generation of services if the policies would match what the technologies are capable of. The elephant in the room is that good Internet will continue to eat the lunch of Cable TV and phone. Bandwidth caps are more about keeping you in line as a "good consumer" than the economics of IP transit.