The only thing preventing it is the content owner's attachment their obsolete business models. I'm sure many people in the media industry realize this, but they earn their bonuses in the present, not in the shiny future. But make no mistake: in the near future, we will have ubiquituous access to most cultural artifacts of the human race, regardless of the business model.
However, delivery is only part of the puzzle. Divvying up the pie is orders of magnitude more complicated. One suggestion might be to take all the money brought in by a subscription system and distribute it based on what percentage of total views each piece of content got.
But... not all content is created equal. Think about it for my software by subscription example. If Angry Birds gets downloaded 100,000 times in a given month and AutoCAD gets downloaded 1,000 times, should Angry Birds get a 100x bigger piece of the subscription revenue? Also, the catalogue would never be truly exhaustive because the writers of ClarisWorks might be impossible to get in touch with, or might not agree to the terms of the marketplace.
When taken in its most general terms, Hollywood's business model is strikingly similar to that of venture capitalists: "spend a lot of up front money to make something now, and hope that it generates even more money down the road". There's nothing Hollywood would love more than to give up their bonuses in the present for a future where content created now will continue to bring in small amounts of money for the rest of their lives.
I probably come off as sounding too sympathetic to Hollywood. SOPA is pretty clearly the wrong answer to a question very few people were asking. But I think we do ourselves a disservice by downplaying the complexity of the problem.
Also, digital preservation is not a one time expense.
I like the idea of a subscription model for a complete catalogue of content too, but it's not really feasible. Imagine if someone said "I want to pay $20/month to be able to download and install all software ever written".