We seem practically bound to have epimemetics in one form or another. However, it seems destined to be a controversial topic. I'll say briefly here what I think it should mean:
Firstly, I'm pretty happy with Waddington's use of the term "Epigenetics" to refer to influences on development not coded for in nuclear genes. I'm not happy with the subsequent hijacking of the term to refer to non-nuclear cellular inheritance. Since genetics is - or should be - the study of heredity, the we can't call the study of non-nuclear inheritance "epigenetics" - that would be an oxymoron. I explain this in more detail here. We just don't need to use a major genetics term to refer to "the study of mitotically and/or meiotically heritable changes in gene function that cannot be explained by changes in DNA sequence". The whole idea is a scientific farce.
That brings us to epimemetics. Using the term in Waddington's manner suggest that epimemetics is the study of epimemesis - or information acquired during ontomemy.
Just as brains acquiring information from their environment during development is epigenesis, so, computers acquiring information from their environment during development is epimemesis.
A complication with the concept of epigenesis is whether "epigenetic" information can consist of other genes. Epimemetics faces a similar issue: whether epimemetically-acquired information can consist of other memes. I favour an affirmative response - but currently I'm happy to leave the resolution of these controversies to future researchers.
Another thing that "epimemetics" doesn't mean is the art of overcoming bad memes.