has several errors.
Here’s the image:
and although I had seen it thousands of times, I hadn’t really noticed it until I saw it for a moment in Lessig’s corruption video. The errors in the Devanagari and Kannada scripts are immediately obvious; the New York Times ran an article about the error (they knew of only the Devanagari and a Japanese error).
Apparently, the original author has lost his source files, and no one knows how to fix it (seriously), so they have been either simply giving up, or using the ingenious argument that the logo is appropriate, as the existence of errors is characteristic of Wikipedia.