The strategy being to first introduce some artificial scarcity on internet bandwidth through data caps, and then offer exceptions. Either the cost to the business to be on the exception list will be high, or the cost to the user will be high to override the data caps, with the cost of service increasing as oligopoly in a region allows.
They can use this to steer people to the services that they back, like Hulu, or at least to make their own cable-box video on demand services seem more reasonable.
If one were truly pessimistic, the categorization of content wouldn't stop at simply high bandwidth services like video. A separate limit could be applied to things like html, with websites indexed as to their type and views counted. Which could result in an online equivalent of Sinclair media's control of local television news as certain channels of information are more accessible because they don't count towards a cap. And anything untrackable like VPN use goes into a more expensive version of the existing bandwidth-based quotas.
I believe it will not come to that last part, but less because I don't believe the telcos would attempt it than that I think it would incite people to legislative action. If it's introduced in slow increments, though, we might get closer to this worst-case scenario than many believe.