http://technology.timesonline.co.uk/tol/news/tech_and_web/article4557935.ece
the fact that there are real elements in the shot detracts from the grandure of that videos statement
I would not call this a revolution in facial animation. It is an evolution of motion capture. This technology is very interesting and has been around a few years. You are essentially capturing model (scan)data, motion and textures at 30 fps.
Basically it is a camera map of an organic subject, textures are video sequence playing on a mesh that deforms in sync to the textures.
As far as this showing up in games. There are lots of limitations. file sizes are huge since you are storing a video in place of textures. The only use for this is in cinematic sequences as every shot is a one-off. So unless your game has just about no story, since you have to capture every single sequence, and games start shipping on blue-ray disks things are still a bit out. The tech needs to become smarter and dynamic, and more efficient compression or recycling of data needs to be there. Still really cool though.