"The examples he gives are all pop-science"

I don't think that's true of The Selfish Gene (and even more so for its successor, The Extended Phenotype, as one can tell just from the title). It seems to me to be a serious work on evolutionary biology.

It's "pop science" in the sense that he's taking academic work and making it palatable to a general audience, minus the rigor and complexity. His sources were books like Fisher's Genetical Theory of Natural Selection and papers by WD Hamilton.

It's not that it isn't serious. Pop science doesn't have to be BS. But it often risks conveying to readers that they have a better grounding than they actually have, which The Selfish Gene has also done. Pop science books are good ways to attract young readers to the sciences and to keep older readers engaged in it, but they are not in and of themselves contributions to the field.

I'm not really up on the history of these ideas, but my impression is that The Selfish Gene was a contribution in the same sense that Einstein's paper on special relativity was a contribution. Neither really changed what one would predict the result of any experiment to be (the Lorentz transformation had already been produced before Einstein, by, err, Lorentz...), but rather in the perspective by which one views the field, which does affect future research.

He's a populariser of science though. No criticism intended by that. His books are still serious works, but the serious intent is to educate a general audience, not be an important document for a specialist audience. It's a story about his theory. And good popularisers of science, which I think he definitely is (or was?) are good storytellers. He takes a basic thesis then uses multiple lyrical examples to illustrate it. As a layman, I loved that book and read it several times, it made me feel like I had an understanding of genetics, and many of its ideas have informed how I think about things now (the additional, if somewhat vague, stuff he added re memetics in particular). But I'd also say that book is an epitome of pop science, and that feeling of understanding of genetics was just the basic concepts clicking. Which is also fine, that's a really good thing!

Edit: to go back to original article: memorising the stuff in detail isn't hugely useful -- it's a useful jumping off point to investigate in more detail, but the books basic ideas are easily understood, that's the point

Agreed. Was it not The Selfish Gene that coined "meme"? Or was that word used before by Dawkins in research papers or similar?

'Meme' came about when Dawkins modified the word 'mimeme' to sound like 'gene'. AFAIK that was first done in that book.