I think it's helpful to point out that insulin resistance does not always lead to overweight. This seems to be a trend recently, and not necessarily a bad one. Right along with Masterjohn's recent post, there's Stephan's most recent post, "Glucose Tolerance in Non-industrial Cultures." The basic idea, and the lesson I think we should take from this, is that a lot of insulin resistance is a natural, good thing. We've talked about a related topic here on paleohacks (see my answer on a recent thread for the relevant links): what Peter calls "physiological insulin resistance." That's what happens when you habitually eat low-carbohydrate, and Stephan in fact uses it to account for a discrepancy in the data in his post. So, again, the basic idea is that insulin resistance can be a natural, good thing. Making this a little more detailed, we can say: a healthy human body (healthy from birth, Stephan stresses) can go into high-glucose mode and deal effectively with a lot of dietary carbohydrate, or a healthy human body can go into low-glucose mode and not deal effectively with a lot of carbohydrate. A few days, or maybe a week, is needed to transition between one and the other. (If you're living in a seasonal climate, or just one in which food sources vary gradually throughout the year, this is probably not such a big deal, because you make the transition and then you stay in the new mode for a while.)
Now, as to whether insulin resistance "causes" obesity or not, there is a sense in which, even if Masterjohn is entirely correct, insulin resistance is more of a "cause" than leptin resistance. This is a fine point in conceptual use, but I think it's important. If leptin resistance always leads to obesity (or prolonged leptin resistance, whatever), then the trouble is that leptin resistance just becomes tautologically identical with weight gain -- just like it doesn't really help to say that weight gain is caused by an excess in calories in over calories out. We all agree that that is the case; what's important is what causes that. So if leptin resistance is just the mechanism by which we get fat, then it doesn't do all that much good to talk about it as the cause. I'm exaggerating a little bit; it may not be the case that leptin resistance is as closely tied to obesity as, say, "lack of blood to the brain" is to "death". But the most important point is that it is more closely tied to it than insulin resistance is. So in fact because insulin resistance does not always lead to obesity, it can actually be considered as a real cause more so than leptin resistance can.
So the really helpful thing about what Masterjohn and Stephan are doing is pointing out that insulin resistance can be a cause of obesity, but it is not always the cause, or it is not always, when working alone, the cause of obesity.
(If you know that insulin resistance does not always lead to leptin resistance, and if you know -- separately -- that leptin resistance is just the mechanism by which we get fat, then this allows you to conclude that insulin resistance does not always, by itself, make you get fat.)
So what about the real meat of Jae's question? I think that in the end, if we still know that excess carbohydrate in general (regardless of kind, so even if it's not gluten or refined), in conjunction with other factors can lead to obesity, then what's the point in risking it? Why get 90% of your calories from sweet potatoes if you run the risk of obesity -- because it's not also the case that you have been ingesting, all your life, limited linoleic acid, no gluten, and no table sugar (to say nothing of vitamin D, etc.). It might also be the case that we'll uncover a culture somewhere else in the world, without any traces of the diseases of civilization, whose members drink soybean oil all day long, but have never touched gluten or sugar, and whose carbohydrate consumption is in the 5%-30% range.
Does this sound familiar? It should. Tolerated is not optimal. Kurt Harris was "so bored with the Kitavans" a long time ago. I think we can still be bored with them. Until we see a really good reason to eat a very high proportion of our calories from carbohydrate, I don't see that all this makes a difference.
Most importantly, the insights of Taubes (or whoever can claim them first) remain. Weight gain and weight loss are still not about consciously manipulating one's input and output of calories (because those are not independent variables, etc.). They are still about altering what's going on with the hormones in your body. This is confirmed by a great set of posts by Stephan from some time back: the "Body Fat Setpoint" series. (Here's the first one; look at the archive from January 2010 for the rest.) Stephan gives us a leptin-centric account of fat gain and fat loss -- but much of it accords with what Taubes has to teach us. Everybody wins.
One more note: I think Masterjohn makes a little mistake in his post. He says that when the liver becomes insulin resistant it will continue to produce glucose from gluconeogenesis, and keep sending glucose out into the blood. That may be the case, but why the complexity? Why does everyone like talking about gluconeogenesis so much? Whatever happened to good old-fashioned glycogenolysis? Breaking down the glycogen that is already there in the liver is surely the first option for the liver; gluconeogenesis comes after, for obvious reasons (because you have to burn protein). I would guess that Masterjohn is used to talking about the situation in an LC or VLC body, where gluconeogenesis is much more common. I think it's a small oversight. (Correct me if I'm missing something.)