I think a good interim solution would be to have a client-side filter, browser extension or such. Wikipedia content is very well categorized, so it's not a classical content filtering problem.
I was curious just how well the categories are exposed to clients. I didn't find any HTTP headers or HTML meta tags, but a text scan for "Category:" would do well.
