[detaro]

Many wikis have open APIs (e.g. mediawiki instances typically do). Not sure about MDNs custom thing, but scraping is always an alternative - or even working off one of the HTML dumps available for offline use (although loosing history and original format sucks).

[brundolf]

Theoretically since you can edit the original-format content from the website, you could scrape it from the website. I'm just wondering if this is something anybody is thinking about/working on.