Open Access News points out that the Open Archives Initiative is seeking public comment on their specification for conveying copyright metadata on Open Access content. Such a specification is critical to the aforementioned goal of assembling course packs out of Open Access materials. Open Access is a necessary but not sufficient condition to create coursepacks. Beyond access, we need permission to reproduce the documents for educational purposes. Anyone who has ever had to wrestle with copyright clearances for an online course knows what a time-consuming (and often expensive) hassle this can be. In many cases, it is prohibitively costly in time and money.
However, with a machine-readable Creative Commons license attached to the document, we can then set up a Furl-like system (or possibly Furl itself) to collect downloadable copies of the documents into student coursepacks without concern for violating the author's rights (not to mention copyright law).
Even with this addition, we'll still need a few more pieces. For example, I understand that there is currently no easy way to embed this kind of metadata in PDF documents. Still, this specification (along with it's widespread use to attach Creative Commons licensing) will be a critical step forward.