What chance is there that MediaWiki will *ever* be useful for this purpose? Its developers are only interested in Wikipedia, and in increasing the power of their sysop power structure by adding on a permission-based model no one really needs, but which makes them feel powerful.
They haven't even cloned the GetWiki facility yet, and it's the ideal way to keep those who fork off the GFDL corpus as close to the core corpus as possible. That is of course because they are trying to stop all other GFDL corpus access providers, and retain trademark power over the name "wikipedia", which is actually generic. There are many wikipedias, and the contributor is not seeking to enable or give their work to any specific bunch of "Wikimedia" thugs, they're seeking to give it to all wikipedias.
We should be more concerned with edits, votes and bets and how answer recommendation might move things from Research Wiki to Publish Wiki. Importing a lot of sysop-approved biased nonsense from Wikipedia should be low on our list of priorities. Why not just use GetWiki and get it from Wikinfo instead?
GetWiki discourages the construction of a proper fork by allowing users to fetch articles from Wikipedia on demand, whenever they access a page which doesn't exist. This means that a large proportion of the content hosted by a GetWiki site is actually controlled by Wikipedians. What's more, Wikimedia would be within its rights to cease service to any GetWiki site, leaving them out in the cold with a useless leech script. Why not just download the database and end your dependence on Wikimedia? -- Tim Starling 11:50, 23 Jun 2004 (EEST)