New Atheist books are a dime-a-dozen these days. Stewart's work is not the best of this type of books, but by no means the worst. To his credit, he is more widely read in philosophy than are most of the writers in the New Atheism school (though that isn't saying much); the overriding fault in general of the genre is those who think that simply being a physicist or a biologist is sufficient to allow one to finally dispose of all the great philosophical questions, such as the nature of truth, free will, morality, and the existence of God. Stewart earns a solid `B' in that regard; he at least has read some philosophical background and tries to enter into the debate at a somewhat more sophisticated level. Still, he seems to believe that you can settle the problem of evil or free will or moral objectivism in just a few pages, thanks to Darwin. But Darwin's theory does not solve philosophical problems; nor does science in general resolve metaphysical problems. Those who think that it does simply end up doing bad philosophy.
Consider his argument against moral objectivism. Stewart believes, following the predictable pattern in these books, that since ethics is the product of evolution, then morality has no objective basis, and we can do whatever we want. Let us leave aside the fact that, contra Stewart, there is no solid evidence that ethics is the product of evolution (at most there is a series of mutually inconsistent, speculative theories about how morality might have evolved). Also leave aside that even Stewart himself admits that morality is only partly a product of evolution, at best. But even if morality were a product of evolution, it is a logical non sequitur to hold that ethics cannot be objective. After all, science is a product of evolution too. Stewart makes an effort to address this argument (again, unlike most writers in this genre), but it is misguided to think that one can resolve this profound philosophical problem by reading a few books and then announcing that the problem has been solved. Further, Stewart can't fully bring himself to accept the implications of this conclusion. He can't admit outright that he is really a moral nihilist; the book is interspersed with claims of nihilism, but also frequent expressions of moral outrage, for instance at the way we treat animals (as well as a half-hearted attempt to demonstrate that utilitarianism is the best moral theory -- which doesn't even make sense, since if morality is subjective, then there is no basis to value one theory over another). But you can't have it both ways: if there is no morality, then there is nothing wrong with the Holocaust or with factory farming. As so often, the attempt to have it both ways gives you the worst of both.
Or consider one of his arguments against religion. Again, to his credit he insists that he won't commit the typical New Atheist tactic of refuting only the most crudely anthropocentric conception of a deity. However, he inexplicably proceeds to do just that, rejecting a more philosophical, non-anthropocentric conception of God because it violates the "original" meaning of the idea of God, which he claims is essentially anthropomorphic. Let's consider this argument.
1. What could "original" possibly mean in this context? Was there a date, time, and place when the idea of God was first established?
2. Even if there were (absurdly) such an "original" meaning, why should it have a claim to be the TRUE meaning? For instance, the original meaning of "atom" is "indivisible" (and in this case we really DO know the original meaning!). But that hardly entails that scientists are currently misusing the term.
3. Stewart adds (126)that the non-anthropomorphic conception of God is not "in all probability" what the "authors of the worlds' sacred texts" believed. Leaving aside what he could possibly mean by "in all probability," or his belief that the "authors" of all the world's diverse sacred texts all shared a single meaning (perhaps they were all present at the "origin" of religion??), still the claim is demonstrably false. In fact, even the most cursory knowledge of world religion reveals that many religions have an abstract, non-personal notion of God: e.g. Taoism, Hinduism, Zen Buddhism. And even in Christianity, there is a long tradition of rejecting crude literalism about God (did he really "walk" in the Garden of Eden?).
As with so many New Atheism books, Stewart has not done his homework in purporting to give us a final refutation of religion, and ends up attacking a straw man (straw God?).