You can tag arbitrary content on the web, you can do it in a low-tech way to make it easy for everyone to do...But...How do you find instances that people haven't tagged? Or deal with overlapping meme labels?
A meme ID is like a magic bullet in the document saying "I am about this meme." Can we construct a general taxonomy of memes, and then specialized lists of meme IDs that authors will feel comfortable adding to their documents?