After being so impressed by The Selfish Gene, I got The Extended Phenotype for Christmas and started reading it recently.
The Selfish Gene talked about scientific concepts but was definitely geared for the lay reader. It wasn't light reading, but anyone with a modicum of knowledge in the sciences should be able to read it without any problem. The Extended Phenotype is rather denser and harder going, with more technical language. There's a glossary but it doesn't change that I have to go over paragraphs a few times and struggle at them, like reading a textbook and trying to understand. Dawkins doesn't hesitate to use words like pleiotropy, or assume that the reader has a basic idea of why haplodiploidism leads inexorably to sex ratios of 3:1; and he frequently refers to things any biologist would know but laypersons might not, without explaining them or with only a cursory explanation that one must struggle with to proceed. It's probably midway between a populist science book and a paper in a peer-reviewed scientific journal.
Like The Selfish Gene it's chock full of amazing and interesting anecdotes about animal behavior which are mentioned along the way to explaining or examining some question of inheritance. For instance, in explaining why an evolutionary "arms race" might lead to the cyclicality of locusts, he also gets to explain why both breeds of locust happen to have a cycle whose number of years is a prime number (13 and 17). His stories of those breeds of ants which take over another hive instead of forming their own, or steal workers from other breeds, lead to some fascinating issues of what evolutionary counters are and are not possible. And his explanation of why lions mate so remarkably often (on average, once every 21 minutes) despite the females not ovulating during most of it (copulation induces ovulation, in fact) makes one re-examine lots of assumptions about evolution.
I'm only about a third of the way through, and he promises the meat of the book won't come until the last chapters; so it would be unfair to say that so far the book feels apologist and unfocused, because that's intentional. I hope I'll feel as the author does that the end makes the earlier parts justified. Dawkins considers this his best work, and while I have often disagreed with an author about their choice of best work, I often find it a better bellwether than popularity. (A great example is Nietzsche, who liked The Gay Science far more than the much-hyped Thus Spake Zarathustra, and I entirely agree with him.)