This was one of the things we worried about when initially designing dotLRN.
We figured dotLRN is used to manage multiple communities, so it's already a
bit of a meta-tool. Mutliple dotLRNs means a meta-meta-tool.
This doesn't mean it can't be done. You'd need to set up a different community
type root for each instance of dotlrn, which the current model would support
without too much difficulty. That's really the biggest requirement - making sure
each tree of communities is rooted independently at each main dotlrn mount
point.