Nominally, this proposal will have no detrimental impact on any current service. Put simply, ISPs are being given the option to offer a "premium" service to those data suppliers who wish for their content to be delivered at a "premium" rate, at a premium price, thereby improving their perceived web experience.
To the simple-minded, this is a perfectly straightforward case of adding value to a service and charging for that added value. Nobody has to pay anything extra if they don't want to. However, this doesn't address the brutal reality.
Firstly, ISPs already saturate their bandwidth as far as they're able in order to be competitive. The creation of an express-lane for premium content will, by default, require the degrading of non-premium content delivery. Certainly the increased revenue could be used to improve infrastructure and have a net benefit on all bandwidth, but ISPs are businesses and it's fundamentally naive to assume this will be the result.
Secondly - and more importantly - this move would change the culture of the web irrevocably. In the first instance, content providers will have to pick a camp, and we will be faced with a two-tier system. Two-tier will just be the beginning though, and companies will have to quickly start incorporating their "content deliver" streaming costs into their business strategy. Like any variable, contracted service, it will be open to competition, abuse and legal dicking-about. It will change the very nature of the web, and we will all suffer from the lack of an even field.
A more subtle problem would be the loss of impetus to improve the efficiency of data delivery. As things stand, it is in every single person's and organisation's interest to constantly strive to improve the bandwidth-efficiency of their sites, languages, algorithms and services. As soon as the big guns find themselves able to take a short-cut to improving their users' web-experience by paying for it, half the major driving force behind these innovations in efficiency will be gone.
I'm sure there are many other reasons to oppose this change, and I honestly can't think of any compelling reason to approve it - unless, as I said, one takes the short-sighted, uninformed (or plain greedy) stance that this would improve certain uses of the web, at least for now.