People sometimes ask, "Hasn't this already been done?" It would seem that it hasn't, which is why so much of the implementing code has to be
designed
and developed rather than borrowed or reverse-engineered. In some ways,
the
closest project to this one may have been been the various proprietary sites that used the Wikimedia update feed servicehttps://meta.wikimedia.org/wiki/Wikimedia_update_feed_serviceto stay continuously up-to-date with Wikipedia, but to my knowledge none of them used MediaWiki as their engine, and their inner workings are a mystery. Those also tended to be read-only rather than mass collaborative sites.
Id say that http://getwiki.net/-GetWiki:1.0 was similar to your "superset" concept (minus the merging part)
there will inevitably arise completely Inclupedia-specific matters that need to be dealt with in a different venue. Presumably, it'll be necessary to create a whole new infrastructure of bug reporting, mailing lists, IRC channels, etc. But, I want to get it right from the beginning, since this is an opportunity to start from scratch (e.g. maybe there is a better code review tool than Gerrit?) I have created Meta-Inclu as a venue for project coordination.
Be careful here - well its important to have bug tracker, etc - concentrating too much on support infrastructure and not enough on the actual issue at hand is a way that new projects sometimes fail. These types of things also tend not to be needed by small just-starting-out projects in the same way that large projects need them. Of course every project is different and you are in the best position to evaluate your project's needs.