Not so fast my friend. Video breaks this assumption especially when you consider increase in broadband penetration rates over next few years. Last mile, which has been the bottleneck for the first 20 years of internet will find that the other parts of the network backbone, peering points etc will need costly upgrades to carry all the video traffic. Especially as the video content on the internet moves from a measly 320x240 largely to HD, 4K and beyond. The increase in resolution and increase the demand for higher resolution (broadband penetration) can possibly change the economics to a physical media based distribution model especially for high resolution video. Even mobile video can consume a lot of bandwidth if it has interactivity. In many ways, relatively low penetration of braodband has been the reason that has kept the internet from collapsing under its owne weight.
Network based content distribution technologies are evolving to help address this. But my point is that the outcome is not as certain as you make it sound.