The "what ifs" people breathlessly talk about when they get into this sort of discussion [about SF vs. mainstream fiction as a literature of ideas] are the premises which drive SF stories, and aren't necessarily the themes they address - assuming the story actually addresses theme at all and doesn't just wallow in empty worldbuilding and pie in the sky speculation. It is true to say that SF is a "literature of ideas" if you take "idea" to mean "premises" - the critical acclaim in which SF stories are held often correlates strongly with how wild and out there their premises get. But there is no story in the world which doesn't offer premises. Science fiction isn't the exclusive literature of ideas; at most, it's the literature of freaky counterfactual premises.
The discussion was in general about SF as a literature of ideas, something I've touched on here before, and how mainstream litfic is no less suited to being the same sort of thing with the right approach.
The ideas in Jean-Paul Sartre's Nausea are no less challenging than the ones in Frank Herbert's Dune; the packaging and delivery is what's most different. I suspect more people are attuned to reading the latter than the former if they must choose.
Take the test yourself: what's more interesting, a guy poking around some French town musing about the nature of his own existence, or a war over a desert planet with gigantic sandworms that secret the spice which drives the universe? (Okay, it depends on your definition of "interesting", of course, but you get the idea. Most people I know would pick Paul Atreides over Roquentin as someone worth reading about, but that doesn't mean Sartre wasted his time.)
What the folks at Ferretbrain touch on in this passage, though, is something I've come back to a lot: just because a story contains something doesn't mean it's about those things. You can confirm this by reading any number of SF stories that contain any number of speculations about technology, but use them as little more than fuel for a story that might not require it in the first place.
A real literature of ideas isn't just a dumping ground, but unfortunately it's easy to satisfy most readers by simply throwing in what passes for ideas by the fistful and expecting everyone else to connect the dots. A scene in which two people talk about a deep philosophical concept is not so much about that concept as it is about the two people talking (c.f., Marguerite Duras's The Square). This isn't to say such a scene can't be used in a story that does have that concept as one of its real subjects, just that such an approach is what passes, most of the time, for such inclusion, because the real thing is far harder to single out.
One side effect of this is how finding the "real" subject of a story moves that much more into the hands of the reader. Again, not to say the author can't have his own intentions about a story's real subject, and that his work to put that into the story won't be honored. But it does mean all that work constitutes a beginning more than an end, a starting point and not a closed book (in more than one sense).
Now I'm going to turn around and suggest something that'll seem wholly out of phase with what I've been talking about: a literature of ideas may well be something appreciated after the fact, not constructed in advance. Meaning it might be better to develop our skills of appreciation for the stories that resonate most strongly with us, rather than trying to shoehorn an excess of ideas into a story and push that as our "literature of ideas". What we call the "literature of ideas" might be best constructed as an oeuvre in retrospect, in big part because of the way the finest writing is constructed in the first place.
Every author embodies his outlook — his ideas — in every story he writes, even the minor ones. (Sometimes it especially is the minor ones, because that's when his guard is most down.) The best authors have the most wide-ranging vision and are able to communicate that vision most effortlessly in their work; they make it seem almost like play (c.f., Dostoevsky, especially in something like Crime and Punishment). We look back on such work and marvel, and then — not always successfully — take that as our model for how to proceed, forgetting all too quickly that we are not Dostoevsky.
A "lesser" author (though by no means a bad one) makes such a project into work rather than play, although the sheer level of effort he puts in often redeems the project anyway (c.f. Thomas Mann, The Magic Mountain). I say "lesser" in quotes because, let's face it, anyone able to write something like The Magic Mountain is a big-league player and not someone to be dismissed lightly. But the manner of integration matters.
New York City
Other Lives Of The Mind