Discourse being an Open Source application itself and dedicated to Open Standards, it theoretically build on already existing interface specifications, that try to federate conversations accross platforms.
The two leading initiatives in this directions are the
As federation is not trivial, I know it from experiments with the Federated Wiki, initiatives like the above will help to find common consensus on how to model distributed conversation graphs. This will likely also include ideas coming from the Open Annotation working group that also works on Hypothes.is.
Further more, as I am interested in aesthetically designing these distributed/federated conversations, the same question has already been raised earlier and could even be used to some weird scenarios as merging two Discourse instances, now that I think of it.
The important part here is to get the content of one site to the other, with possibilities of Forking and Re-Merging, i.e. Pull Requests known from Git(Hub).
One of the noisy results when searching for quotes that I provided delivers another small piece of the puzzle. In the end, the user’s interface itself should facilitate the creation, visualization and maintenance of distributed conversations. This is a task not to be solved by Discourse alone, as mentionned earlier.
But it can help with that, as its name implies sociological reasoning about the underlying principles, regulated algorithmically by its codebase.
Just my € 0,02.