For mainstream music consumption, streaming music is the future. Anyone still expecting to build a business selling a la carte downloads at this point is laughably out of touch. There are few blogs out their that are written by jaded ex music industry guys that just can't get over the mental hump regarding streaming music. These aren't the people that you should be taking advice from when it comes to technical paradigm shifts.

There have been no successful implementations? Wow. Spotify has been around in the US for what, 2 years? A little early for the funeral, isn't it?

There are really only two problems that need to be solved, and both are quite solvable in time:

1. In order for streaming to work, it really needs to be free (or "free like" to consumers) AND subsidized by a major company. The combination of a subsidy and ad revenue should be enough to push per track income above $0.01, which is really the point at which this model starts to become economical. Yeah, I know it takes 100 streams to equal one iTunes download, but when you've got 10x the transactions BECAUSE IT'S FREE, and people listen over and over again (assuming you write music that sticks) those pennies will add up.

2. The companies that will eventually subsidize streaming music (ISPs, Cell Companies, Cable companies) are not able to do it yet due to fact that the bandwidth generated will cost them too much money. Give it about 5-10 years and you will see the cost of bandwidth drop the point that it becomes economically and technically feasible for Sprint/Verizon/AT&T/Comcast/Telstra/etc. to figure out that offering free streaming music is relatively cheap competitive advantage.

Sure, press vinyl, CDs and cater to audiophiles/dinosaurs/hardcore fans/different tiers/preferences, but the big 800 lb gorilla is going to be streaming.