The closing question is what interests me : the question soon will not be “how do we scan this?” The question will be, “What do we do with all of this scanned, real world spatial data?”
I see one of the break throughs being with object recognition within these points clouds and then attaching data to these clusters, negating the need for modelling at all in many situations. Shape and pattern recognition directly feeding machines for reproduction and replacement as well as using micro cameras to monitor and model objects in motion such as motors and vibration at the microscopic level. Imagine replacing all cameras, SLR, mobile and cctv with this technology. Public domain data sets of the existing world as it is now, not yesterday or last month but RIGHT NOW!