Dropbox makes a profit (just barely). Google is still dodges investor questions on YouTube profitability which suggests it doesn't. Flickr/Soundcloud have been fighting for survival for years now.
The power law distribution is problematic because unless you can predict which content becomes a hit, you either treat all the content like it would (extremely expensive), as if it wouldn't (extremely slow and still expensive), or some smart predictive in-between (which is a very hard problem that you can see Netflix/HBO and even Youtube suffers from). You can see Dropbox is working on smart-sync too, because it's pretty critical for keeping costs down.
Besides, you really don't want things to go viral. Viral is expensive to serve and expensive to engineer for.
I think torrent-style distribution is a brilliant idea. Not sure how people will feel about their phone data or metered connections being used, nevermind the engineering challenges of getting torrenting working on a cellular connection.
You're asking Dropbox to essentially become Google without the profitable part (search ads). Way easier said than done. There's very little profit to be made from being a commodity, it's not clear that becoming an umbrella of multiple commodities is any different.
Dropbox is certainly positioned to make a good product, so is Google/Microsoft/Amazon/Apple. But until people actually start paying for good products so they're profitable products, they won't be built or will keep disappearing to be replaced by subsidized/free mediocre products by Google.