I've been trying out the David SLS2 system as well, although I got it all from the same source. I found that while the initial scan has a lot of resolution, that seems to degrade as scan gets added to scan to build up a solid model. Since it only captures a single perspective at a time, it's necessary to do that in order to come up with a model you can print, and the process of combining scans is not altogether easy, especially if ones model doesn't have a lot of readily identifiable features to match point-by-point. Has any of this been an issue for you? Have you figured out any work-arounds?

Andrew Werby
Juxtamorph.com