Another starting place is looking at the IMS (
imsproject.org) specifications for the XML dictionary. Specifically the content packaging and meta data specifications are among the more frequently implemented and probably applicable here.
I realize that full "compliance" with these specs is probably well outside your scope, but if we all work towards IMS I think we can maximize code reuse and minimize duplication of effort.
OCW (ocw.mit.edu) announced at the EduCause conference that they intend to use these specifications to transfer data between the various MIT course management platforms and OCW. This means that eventually SloanSpace will implement a subsection IMS. OCW is still in the process of creating technical specs, so I do not yet know the time frame and I certainly can not guarantee that complete "compliance" will be in scope for Sloan either.
That said, my long term vision is that if indivdual dotLRN projects implement peices of IMS specifications to meet thier needs, that one of the OACS companies will find it in thier business interests to take it the remaining delta to compliance.
Michael Feldstein is another big IMS fan and a bit further up the learning curve on IMS then myself. I'm sure that he, as well as myself, will be happy to help you in interpreting how the specs can apply to FAQ.