Here's a video where Eric Stollnitz from Microsoft Research explains the steps behind generating what we know today as Photosynth 2 spins:
By way of source, the Preview's 'About' page http://photosynth.net/preview/about/ says,
"Third, the technology uses the feature points in each photo to generate 3D shapes. It does so on a per-photo basis rather than trying to generate a global 3D model for the scene."
I know that's a lot of links and perhaps more information than you wanted, but hopefully it gives you a sense of where to keep an eye on (SynthExport), who to talk to (Christoph, Josh, and Greg), and source code to look at (SynthExport is on CodePlex and PhotosynthToolkit is on Github).