This looks like it hasn't advanced to the point where photogrammetry already done by http://realities.io/ is today. They don't need many cameras in parallel, they do their work with a single ordinary DSLR. And they already fuel actual VR applications with their data.
This approach is limiting.
Approach described in original post is much more universal and useful for real things.
It looks like you're new here. If you want to get involved, click one of these buttons!