The potential is there. Via Slashdot...

http://news-service.stanford.edu/news/2008/march19/camera%20-031908.html

It's not at all clear how this differs from current sensors with microlenses like the Sony designs.
As far as I can see, you can do this today in software taking a 10MP sensor and reducing it to 2.5MP using groups of 4 pixels. There too you will get groups of 4 pixels with slightly differing viewpoints. In fact with the Bayer filter it's already doing that.
S/N ratio of these tiny photosensors will be poor, so I don't see it replacing current systems for years. But it is an interesting application.

Makes me wonder I wonder if anyone's patented my idea yet?
Must go check ... no I'm not going to tell you what it is ;)

It's not at all clear how this differs from current sensors with microlenses like the Sony designs.
As far as I can see, you can do this today in software taking a 10MP sensor and reducing it to 2.5MP using groups of 4 pixels. There too you will get groups of 4 pixels with slightly differing viewpoints. In fact with the Bayer filter it's already doing that.
S/N ratio of these tiny photosensors will be poor, so I don't see it replacing current systems for years. But it is an interesting application.

Makes me wonder I wonder if anyone's patented my idea yet?
Must go check ... no I'm not going to tell you what it is ;)

There's always going to be some distance between different pixels, but since a single lens handles all this before the light strikes the sensors, I wonder if their design that uses separate lenses for each may be superior for true 3-D effect?

I guess we'll have to buy millions of lenses now, not just single ones :D

Check out http://www.freepatentsonline.com/ btw

I am not going to change 10 million lenses from wide angle to telephoto !