The movie above provides a glimpse of the output which interestingly is not dependent on LiDAR. C3 capture the data using calibrated cameras that for every picture knows the positions at angles of capture at a high precision. This enables each pixel to have a known geographical position with decimeter accuracy and thus using a combination of images a 3D model can be produced.
The company has just released their Oslo model and it is well worth taking a look at.
See http://kart.sesam.no/3d/ for a 3D flythough (Windows only at the moment) and http://www.c3technologies.com for more info.
Thanks go to Frank at the Google Earth blog for the email prompt to put up a post on it :)