Wikipedia is a tempting redesign project because many designers can't seem to understand how a website can be both ugly and highly useable.

I'm curious: why is Wikipedia in need of a redesign? George's reasoning here is that, "It hasn’t been changed or [refined] during the last 10 years. The web and its technologies has developed further and so [have] users." I'd argue that this puts the horse before the cart, though: Wikipedia's design shouldn't change simply to make use of new technologies unless using those technologies could make the site easier to use.

Figuring out the issues with Wikipedia's existing design are the real challenge: it's ugly, sure, but where does it fail functionally? There's a great article from 2007 in which a designer lays out where Wikipedia fails in terms of usability: http://www.raizlabs.com/graiz/2007/10/21/on-wikipedias-design-or-lack-thereof/

Some of the problems the designer identified in 2007 have been addressed, and others haven't, but it's interesting that the problems he identified are largely not very sexy: they really just involve moving navigation controls around, finding a better way to construct breadcrumb trails to enable users to locate themselves within the encyclopaedia etc.

I'd like to see a redesign in which the goal was to keep Wikipedia pretty damn ugly while at the same time making it more useable. That'd be a great challenge.