Mark Pilgrim has an interesting take on the Creative Common’s license. The CC license is a copyleft framework that allows content creators to choose what rights to retain, and what to grant explicitly, rather than the all assuming copyright ©. The GPL was designed for source code. Applying it to other media runs the gamut of success stories. CC was intended for all possible works… from written text, images, video… anything that you would traditionally stick a © on.
A encapsulated issue here is something archive.org was created to handle. The nature of all things digital — especially things residing on the internet — means everything is susceptible to dramatic flux at the will of individuals, ISPs, general neglect, etc. If “Source,” once given freely can be so easily removed without consequence or enforcement, it really means the freedoms granted by CC are asterisked, doesn’t it? There is a catch. Without the aid of organizations like archive.org or great care taken by the original creator, “digital source” is transient.
GPL’d source code I don’t believe has the same issue (or at least not nearly the magnitude). You have SourceForge, Linux package distributions (Debian, Ubuntu, Fedora), etc. who are set up to retain and distribute the source code as well as the binary it created. There is basically an institution and infrastructure in place to protect the “digital source.”
Taking a step back copyleft is fundamentally a new thought process. CC is much newer and much less prevalent than GPL. CC is a wonderful thing (especially when you can search databases for things (pictures, video, text) based on the type of CC license you need), but I think we will have to wait and see how the “remix and return” mentality holds up.