How exactly should these images be treated? I'm talking about the image tag http://en.wikipedia.org/wiki/Template:Wikipedia-screenshot and it's category http://en.wikipedia.org/wiki/Category:Screenshots_of_Wikipedia. The template only says that Wikipedia text is licensed under GFDL and that Wikipedia is copyright of Wikimedia, but it doesn't rely address the issue of the copyright status for the image itself.
According to http://en.wikipedia.org/wiki/Wikipedia:Image_copyright_tags it's a "fair use" tag (that's where it's listed anyway), but we are not currently treating it as such. Gmaxwell's bot originally tagged these images as "far use orphans" if they where not used in articles, but after a storm of protests (well 3-4 anyway, I raised some questions myself back then) he (reluctantly) excluded them from his bot. Problem is you can't rely argue that they are free images, as many of them include either copyrighted GUI elements from the users browser and OS, and a lot of others include copyrighted "fair use" images from articles.
The way I see it we have 3 choices.
1) Keep using it as a semi-free "special" case, maybe we can argue that the screenshot as a whole give enough context to declare the inclusion of copyrighted images to be within fair use or some such (IANAL).
2) Make it a GFDL tag, and run an extensive cleanup project to crop, blur or otherwise remove all copyrighted elements from the screenshots (and clean out other mis-tagged junk).
3) Confirm that it's a fair use tag, and delete all the images that have not been used in any articles for at least 7 days (in other words most, if not all of them).
Any thoughts? Whatever status they should have needs to be made more clear IMHO.