Just a very quick note on a moment of insight--
I was reading about schema theory (as discussed by Doug Brent, and also Richard Anderson, but as is rooted in Piaget) for another project, and all of a sudden it really clicked with Shaughnessy. Anderson suggests that teachers often view mistakes as mere "blemishes" when in fact they may be evidence that the student posesses a entirely different schemata.
I'm still not entirely sure what Shaughnessy's approach would look like in concrete practice, but it seems to me it would mandate a lot of conversation as a component of draft reviews.