From my personal experience, I occasionally try to see some stuff from the Youtube channels, but for one really good video there’s lots of stuff that really doesn’t appeal to me.
And I can’t really say nothing about Netflix, since it is not available in Portugal.
There is already lots of good independent production out there, amateur or professional. I just don’t know how and where to find it.
The solution, thus, is simple:
* Develop a boxee like setup box for around 50$/50€ with an internet connection;
* Distribute it around the world, no region locking whatsoever;
* Have a really good recommendation engine, based on social or whatever. It should allow you to input ‘titles’ that may not be available on the platform, but work as a ‘preference indicator’ (for example, ‘I like House and Battlestar Galactica’)
* It should allow for social subtitling, perhaps even with a reward system - for free;
* Charge 1€/1$ or less per paid TV episode, 2€ or less for film. Better yet, 10€/$ flatrate per device. But make the device still useful for anyone not willing to pay for that, with lots of properly selected free content;
* Commercials are acceptable: one brake per half an hour, not more than one minute : for non-paying customers.
This is not something terribly innovative (except for the recommendation system), but properly implemented it could be quite disruptive.
I've been on Netflix for about 6 months now, and I'm flabbergasted at the quality of their recommendation engine (which maybe owes much to their Netflix Challenge -- I don't know how efficient it was before).
For every show listed, they predict how much I will like it, and their prediction is almost always right.
On the other hand, Amazon is almost always wrong, and so is Youtube; my guess is Amazon bases its recommendations on what I buy instead of what I like, and Youtube does the same (what I watch instead of what I actually like).
A recommendation engine for Youtube (or all video sites) that would actually work would have a lot of value for users.
There's a chicken-and-egg problem: how do you get enough ratings to be able to make recommendations; this can be addressed by starting in a small niche of specific content: start with a small community of people passionate about a subject.
My two cents.