I have the sneaking suspicion that this study was done and reported internally within Canonical in the days prior to the open discussion on the lists about Unity vs Gnome2 for Natty, but they deliberately withheld it from the public because they knew that the results would damn Unity in the eyes of the public and make it very difficult to proceed with Unity in Natty.
Of course, I could be wrong, and maybe they literally just finished the study mere days after the closing of the discussion on the lists. This would just be a matter of unfortunate timing. I can understand the desire to conduct a UX study using relatively stable software beneath (to get useful results, you know), but if you conduct it so late in the ballgame that its worth is greatly diminished, then why do it?
Unless those crashers are 100% reproducible on all systems and completely obvious, chances are good that they won't be fixed before release. Two or three weeks is not a long time in software engineering. Not to mention that quite a lot of enhancement requests could plausibly be filed in response to the data gathered from the study, and without those enhancements Unity would seem quite "rough", and you see where I'm going with this.