Streaming P2P could (and should) be the future of content broadcasting

BitTorrent's inventor, Bram Cohen, has been talking about new streaming protocols for BitTorrent for a while, and the technology is starting to come around (see video below). Release is expected around BitTorrent's tenth birthday this summer, and assuredly, the media and content worlds will go apeshit over it as it is definitely a new opportunity for piracy. But once again, they'll fail to see how they could use this to their own benefit.






Content-rich streaming websites like Youtube, Spotify, Deezer... have some huge bandwidth costs to "upload" content to your computer. I don't have many exact figures, but I did find this Slate article from 2009 about Youtube, which cites bandwidth costs of $360 million/year (out of $700mil total costs, and vs. $470mil losses). So around half of their costs are just bandwidth, and a huge chunk of their losses too.

I'm not a tech buff, I don't get code, but it seems like Cohen's project is geared more towards specific broadcasting rather than constant streaming/sharing of large volumes of catalog always on-line. But even for live or almost live broadcasts, imagine how much content producers would save by sharing at least part of their bandwidth with the public, and therefore the ISPs.
In France, les "ayants-droits" (everyone from producers, artists to distibutors, those who "have rights") have been lobbying to tax ISPs for a while and the recent VAT hike on DSL subscriptions is a testament to that, but why tax if you can unload, unload that bandwidth onto them. The consumer will probably pay more for their  broadband in the end, but they are already since subscriptions are going from 30 to 35€ this year.
Of course this is a little utopian, but broadband has become a public good, and more and more I think it should be considered as a public utility like water or electricity, and be shared more or less evenly. This would spread out the costs of bandwidth, making it easier for new content distributors to appear, with one less overhead problem to deal with.