I really appreciate the work and research behind this project. As others have noted, the defended binary criteria approach is particularly interesting.
I haven't looked at the code because it wouldn't be useful in our school, but I wanted to mention that I'm interested in a completely generic peer-review system for Moodle. If anything, this would probably become a gradebook plugin in 2.0 or beyond.
The aspects of this that would differ from your approach would include:
* any grade item could be peer reviewed
* any item you've completed could then be peer-reviewable by you
* this could work well for completely asynchronous courses (because students can review work from peers who have already completed the course)
* the instructor can have options such as:
1. item is peer-reviewable or not
2. peer reviewer is anonymous or not
3. peer reviewee is anonymous or not
4. your review of your peers counts for your grade on the item (by what %)
5. your peers' reviews of you impact your grade on the item (by what %)
6. the teacher's evaluation of your review of your peers counts for your grade on the item (by what %)
7. is there a deadline for peer review? when?
I say that your project wouldn't be useable in our school for the same reason the workshop module isn't. We're completely asynchronous, and we have lots of courses with small numbers (5) of students going through them at their own paces. So we never have deadlines of any kind, and it's almost never the case that more than a couple students are working on the same assessment at the same time.
One problem with my idea is that if there is no deadline (as in our school), peer review of students who are gone is one-way and rather stale...but I haven't thought yet of any better ways to approximate peer interaction or collaboration in an asynchronous environment like ours. (We're using forums, but those are just as one-way and stale IMO
I'm not hijacking your thread -- I just wanted to mention the ideas of making peer review completely generic and/or potentially asynchronous. If it's interesting, let's carry on the discussion elsewhere. If it's not, this post can peacefully and quietly die of neglect.