... inevitable rise of Internet video ...
Internet video is more of a direct threat to video rental services, but not so much a threat to most live TV viewing.
Network video will definitely enhance our "on demand" choice and entertainment options for years to come, but it's not a replacement technology -- the same way e-mail didn't replace physical mail, and text messages / instant messaging didn't replace phone conversations. Networks are designed for unique point to point communication -- while all of us are going to the same Broadband Reports website, we all see slightly different pages based on how we're logged in. The content is specifically tailored to each one of us. In general, it only makes sense to use the network to distribute standardized content if cheaper methods don't exist. For example, if you didn't get your operating system updates over the network, they'd need to ship them out to you on physical media at a much greater cost. In the case of broadcast TV, a vast infrastructure already exists for real-time digital delivery. Moving that over to an IP network only makes it more expensive.
It has been my experience that for any technology to succeed it need to either:
1) Make things cheaper, like Voice over IP which leverages the unique point-to-point conversation strengths of an IP network and bypasses PUC-regulated fees thanks to the FCC allowing it to be classified as an "Entertainment Service."
2) Make things better enough that it's worth paying more, like with HD TV. (and by extension, HD-DVD / Blu-Ray) People have been willing to pay more to buy HD TVs, and pay additional fees for HD service on top of their regular subscription service because of the improved experience.
Simply taking the channels you receive today and moving them over to IP delivery doesn't do either of those things. At least not with the network technology we know is on the horizon for the next 5-10 years.
The Internet only supports unicast traffic flows today, so that means if multiple users start watching a video program at exactly the same time, the video server still needs to send out separate video packets to each end-station's IP address. That means that any online video provider needs to add incremental resources for every single simultaneous viewer.
There are ways to get one packet to replicate to multiple end-points by using IP multicast, but it requires specific network configuration that is generally only supported within a single company's network. AT&T, for example, is using multicast to distribute their video streams for U-Verse but the multicast traffic never leaves AT&T's network.
That's highly unlikely to change anytime soon. From a carrier perspective, it presents a huge engineering challenge when a customer can inject a 6mbps flow at one point of the network and that means that 6mbps of traffic now gets distributed across the global/regional network and becomes a 6mbps flow out of every single egress link. So the only way to really engineer for that traffic is if the originator of the stream is also the operator responsible for engineering the network -- and now you're back to the same scenario you have with Dish/DirecTV/Comcast/etc with a content packager getting you content from the media companies.
If you look at the last decade of financial statements, the profit margins for companies like DirecTV/Dish/Cable/etc are going down while the cost of the service has been going up. (Yes, profit in dollars is going up because revenue is constantly increasing, but the percentage of net income relative to revenue is decreasing) It is the producers of the content who are pushing the costs higher, and that is driven by a number of factors that aren't all directly related to greed. The shows we're getting on TV today have a significantly higher production value than shows 10-15 years ago, and the push to re-fit all the cameras / film processing gear to be able to process HD video hasn't been cheap. Even little details like having 5.1 audio soundtracks where TV shows previously only had 2 channel audio adds extra costs to production that simply weren't there before.
I also firmly believe that eliminating the middle man in this case is actually a bad thing because the content isn't a commodity product. Each of these networks have exclusive rights to broadcast certain content, so if you want to watch playoff baseball, for instance, you're going to need to have content from TBS. So right now you have Dish/DirecTV/Comcast/TimeWarner negotiating with TBS and saying "we want this price, if we don't get it we'll remove the channel for our 15+ mil subscribers and you'll lose your advertising revenue." We see disputes like this take place all the time, such as the one that is currently taking place between Cablevision and FOX. The satellite and cable companies have motivation to negotiate these contracts because they are still trying to compete with each other based on price. The media companies still need to participate in this process because the sat and cable companies represent a collective buying group of millions of subscribers.
You take that bargaining structure away, and now there is very little to keep the price in check as long as the channel makes its revenue target. The price will naturally escalate up to as much as the market will allow, which I honestly believe will be much, much higher than folks are accustomed to paying for certain channels.
You also have the problem of the cost of Internet bandwidth. Median usage today is somewhere in the neighborhood of 4GB/mo for data consumption. Broadband networks can absolutely be scaled up to support vastly more bandwidth than that, but it's going to take a lot of infrastructure investment to make that happen. It's not unrealistic that your $50/mo broadband connection could end up costing you $150+/mo to have an infrastructure built out to support video, and that cost doesn't even include the cost of the content which is now a separate charge.
The Internet doesn't magically make everything cheaper -- while bandwidth does have the potential to be nearly unlimited, at any given point in time there are very real capacity constraints, and the costs to increase that capacity are not insignificant. That's going to temper what future strategies make sense when it comes to network-based video.