I thought there would be more yes, you're so right, so bright, I'm such an idiot for not seeing it that way before, but instead I'm finding i) things I already know, and ii) things I don't necessarily agree with.
Really, he's such a well-respected critic, I'm surprised I have the guts to stand up to him (in my head, anyway). It turns out I don't see fiction quite the same way Mr Wood does, and rather than feeling stupid about it, I'm giving myself a pat on the back for having, over the last few years, learned, all by myself, and with the help of the likes of you, how to read fiction, and reasonably well.
The first bit is about narration and perspective, and the tension between the author's point of view and a given character's (and he calls this irony), and how you can tell a word belongs to one or the other, and what it tells us about one or the other. So far, so sensible.
It becomes clear, though, that's he's cherry-picking his examples to serve his point.
I'm not here to stage a defense of David Foster Wallace (I'm not qualified) — plenty of others have done that before me. But.
Wood excerpts a passage from "The Surfing Channel" (found in Oblivion). Wood has no problem with Dreiser and Sinclair Lewis (neither of whom I've read) making use of ads and business letters, the "media-speak" of their day, in their work. But "In Wallace's case, the language of his unidentified narration is hideously ugly, and rather painful for more than a page or two" . Ugly. The same tension he's trying to demonstrate, the blurring between author and character, exists beautifully here, but to Wood the words are ugly. To be fair, it seems he's not blaming DFW, it's the language of the times he lives in that's at fault. But c'mon. Ugly. That's a fat 4-letter load of judgement.
Wood talks about verb tenses and time signatures — how things happen at different speeds, now or continuously — detail and randomness, and gives a lovely example of how memory works. Then he seems to contradict himself:
The artifice lies in the selection of detail. In life, we can swivel our heads and eyes, but in fact we are like helpless cameras. We have a wide lends, and must take in whatever comes before us. Our memory selects for us, but not much like the way literary narrative selects. Our memories are aesthetically untalented. 
I think my memory's pretty talented actually. All it wants to do is force a narrative on itself. I'd say my sense of aesthetics evolves from how I process my reality (I have no sense of aesthetics before I'm there to experience it), not the other way round. Increasingly it seems Wood aspires to some 19th-century ideal; Madame Bovary, c'est lui.
So I reached for my copy of Oblivion. And guided by the thoughts I'd jotted down when I first read it, I found:
I am still far from being certain of what the rapid flash of the Father's transfigured face was meant to mean, not why it remains so vivid in my memory of our courtship. I think it can only be the incongruous, near instantaneous quality of its appearance, the utter peripheralness of it. For it is true that the most vivid and enduring occurrences in our lives are often those that occur at the periphery of our awareness.
So while Wood is going on about seemingly random writerly detail intending to denote the real but in fact signifying it, I'm seeing that I favour DFW telling me the actual real, in all its excruciating, unadulterated glory.
I suspect it would take less time for me to actually do my taxes than to read DFW writing about the process of completing one's taxes. But Wood, I think, is looking to fiction for a kind of abridgement of life. Perhaps Wood has trouble sifting through his memory of real-life experience to cobble together a narrative (perhaps most people do?); he's not a writer, after all. Perhaps this is why we read, so others can string together stories for us.
Wood goes on to address character, whether "flat" or "rounded."
Is it contradictory to have defended the flatness of characters while simultaneously arguing that the novel has become a more sophisticated analyst of deep, self-divided characters? No, if one resists both Forster's idea of flatness (flatness is more interesting than he makes it out to be) and his idea of roundness (roundness is more complicated then he makes it out to be). In both cases, subtlety of analysis is what is important. 
I think I agree, though I suspect each reader will gauge "subtlety" differently.
Once character is firmly in place, chronology is no longer so important. Character seems also to be the basis of morality, or at least the framework across which to stretch the moral fabric of life in all its complexity.
The book gets a bit weaker as it goes along. Or maybe I'm just losing interest.
"One way to tell slick genre prose from really interesting writing is to look, in the former case, for the absence of different registers. ."
I'm glad to read this; it sounds like good advice. With this criteria I can confirm that George R.R. Martin's Song of Ice and Fire series has some "literary" merit; it's not just fantasy.
Wood talks about language, word choice, but this is tricky stuff to pin down. Good writing at the word level means something like free of cliché, original, choosing very specific words for effect. Wood has a sense of perfect sentences and gives a few examples, but it seems no rules can be set down. We slide from here into style and metaphor. Wood continues to cite passages, from Woolf, from Roth, and he picks them apart to show how they create the effect they do. Somehow all these things must magically fit together, they must be appropriate and relate to each other (choice of words, metaphor, character, fictional world); there are no specific rules for this (and we should all be glad for that), and somehow we just know when it doesn't work.
Which, all in all, is a very sensible approach to reading. Even though it confirms to me my own abilities as a reader, I'm a little disappointed that Wood didn't reveal the grand secret of good fiction (though, of course, I'd be outraged if he had).
I should note also that this book is not about how to write. I don't think you could derive a system of mechanics from what Wood has set down here. I expect most authors are aware of all these elements, but I wonder to what degree they actively consider them while writing.
Somewhere in this book it was also considered how little control an author ultimately has over the reader, and how much the reader must also act as author in ascribing meaning to the words on the page.
The last chapter is a bit ridiculous. Amid a bunch of nonsense, it's mostly a paean to realism. It becomes clear to me that this book should've been called How (Fictional) Realism Works.
The very best part about having read this book is that I learned something about Saramago's Ricardo Reis, one of the few novels of his that I haven't read yet, about its relation to Fernando Pessoa, whose Book of Disquiet I have queued up. I have a feeling I'll be reaching for Ricardo Reis sooner than I'd expected to.