In my first post I tried to present the case for why memetics is a field worth studying. However, like any infant branch of human inquiry, it faces quite a few challenges. Rather than leaving these as elephants in the room, I'd like to spend a few words clearing the air. If we're going to try and proceed down the path of understanding memetic evolution, we have to be honest with ourselves about what we can and can't expect to achieve.
The good news is that if we define memetics broadly (as I attempted to do), it's really only a small step away from very well-established schools of scientific thought. In fact it's hardly even a new theory; one could easily call it merely a different lens through which to view the fields of anthropology, sociology, evolutionary psychology, and neuroscience. All of these fields, in some form or another, study the structure of human brains and how they change and interact. Memetics only suggests that we think of these interactions in terms of the patterns being copied, rather than the organisms hosting them.
The bad news is that while this lens feels very satisfying and intuitive, it's difficult to apply any sort of scientific method to its investigation. For starters, attempts to design experiments to validate memetics seem quite problematic. There are massive practical challenges in trying to control for all the variables that are inevitably involved in the transmission of memes. You would somehow have to account for the precise mental/emotional state of every participant in the study, an all-but-impossible feat. Beyond that, there are non-trivial ethical problems with trying to manipulate another person's ideas, even if only for scientific inquiry. Given these challenges, it seems optimistic to hope for a 21st century Mendel to come along and demonstrate memetic heredity.
Another method which often comes up is to point at historical or contemporary ideas or beliefs, analyze them from a "memetic perspective," and show how the observed result fits the prediction. While this isn't a completely useless approach, it is one to treat with extreme skepticism. Such hindsight predictions can very easily over-simplify the conditions to result in a neat story, while ignoring equally (or more) plausible arguments which would have predicted an opposite outcome. It's the responsibility of the memetics community to hold this type of argument to a high standard, and call shenanigans where appropriate.
All that said, I do believe that looking at the world through the memetic lens can yield some very interesting observations. After I started thinking memetically, many aspects of human civilization that otherwise appeared inexplicable suddenly seemed fairly obvious. Even applying the high standard I called for above, I believe we can identify some cases where a memetic model really does bring some novel explanatory power to the table. Of course, one of these by itself doesn't prove anything; neither do two or five or ten. But if we continue finding ways in which the model helps us understand reality, then sooner or later we'll have convinced ourselves that it's true.
As always, pioneering new science takes patience and perseverance. It's going to feel like we're groping in the dark for a while, but the rewards of finding the light switch are substantial indeed. So, let's forge ahead!