Dawkins' issue with basically natural vs. artificial selection of memes seems unnecessary as far as 'killing' the meme. There's interesting information there regardless of origin and methods of evolution - a shared lexicon is important (tangentially, in part that's what memes - Internet and otherwise - signify), when discussing memes I do try to distinguish via image memes that are passed on the Internet, classical memes, copypasta, etc.

It's really interesting research, and I don't think it's dead. Me and others built Danqex (formerly Nasdanq) partially to analyze trends and gather information to figure out the lifecycle of an Internet meme. I still also hold out hope that more formal research will continue in this area, whether academic or commercial.