Marking has been surprisingly easy and quick using a simple set of criteria we have developed. One reason I like this assessment is because it promotes engagement with current science and doesn't pander to the lowest common denominator - the high fliers have something to get their teeth into (which is important in the first year curriculum).
In reality, the mark distribution is bimodal, something we've come to expect with our student cohort (particularly in year 1). However, I like this assessment because it's easy to see the students who are engaged. In our last minute, modularized, chickennuggetized educational system, it's difficult for most students to immerse themselves in a task which requires ongoing commitment over weeks or months. Since the pattern of sharing is recorded, there is no place to hide. Even the boffs feel obliged to chuck in a last minute burst of activity, but this doesn't skew the marking at all since the full history of the assessment is transparent.
I'll almost be sorry to see this assessment go next year, but the ever more ponderous Google account admin means that we plan to dump it in favour of Friendfeed (subject to satisfactory trial next term) - where the history of engagement is similarly exposed.