Going back to the really early threads on this topic (pre-dating even the dotLRN bboard), we consistently made a distinction between technical and non-technical "users" (i.e., developers and end users). I think this distinction is wise. From a developer's perspective, dotLRN is (I think) basically an OpenACS application. The dotLRN brand is important mainly so that non-technical users don't have to worry about what's under the hood.
IMHO, the best argument for separating the two sites is that the best advice technical users can get will come from OpenACS developers, who all hang out here.
Now, there are arguments on both sides of this, and I certainly am worried about confusing people or, even worse, inhibiting conversations between technical and non-technical users by segregating them. But we can mitigate that by cross-linking the two sites and crafting our message carefully on both. On balance, I think a non-technical users' site hosted by MIT and a technical users' site hosted by OpenACS is probably the best compromise.
