I recently developed a hypothesis to explain the need for low-carbers to eat some carbs to prepare to pass a OGTT. This hypothesis happens to be quite pertinent to your questions as well, because it basically explains what insulin resistance really is. I wanted to make a blog post about it, but a comment in your thread is as good a place as any to post.

First, insulin resistance does not exist, at least not as it is popularly described, i.e. cells do not respond to insulin, therefore blood glucose remains higher than otherwise. Fact is, cells respond exquisitely to insulin - the liver takes up the bulk of dietary glucose as can be seen by a proportionately small amount of lingering blood glucose after a meal. Consider that in a OGTT for example, it's a bolus of 75 grams of glucose, while the entire bloodstream contains about 5 grams, and this bolus will cause BG to rise only about 3-fold (from 100mg/dl to 300mg/dl), rather than a mathematically expected 15-fold (75 / 5 = 15, or 1,500mg/dl, but at this BG level, we're dying or dead). Even if we test positive for insulin resistance, BG will never rise anywhere near that. Maybe it will go up to 350mg/dl, maybe 400, but not at a level where we're literally dying on the spot. In fact, the BG level for a diagnosis of insulin resistance is between 120 and 140 - 2 hours after a OGTT. As we can see, there is no such thing as insulin resistance.

So, the reason low-carbers need to eat carbs to prepare to pass a OGTT is the enzyme called insulin-degrading enzyme (IDE). This enzyme is mostly located/produced/active in the liver and the brain. The liver is most pertinent for our purpose. Unlike hormones such as insulin for example, IDE cannot be produced in huge relative quantities. With insulin for example, blood level can rise several-fold, while IDE can only rise a small amount. That's because IDE stays within the cells, it doesn't get out of the cells and in the bloodstream - there's only so much space inside cells.

While IDE can't be produced in large quantities, its total quantity can be lowered, that's how eating carbs allows low-carbers to pass a OGTT. If IDE is lower than otherwise, insulin in the blood stays higher than otherwise. As simple as that. IDE level in the liver cells drop because IDE also gets degraded as it performs its various function, i.e. inhibiting ketogenesis, inhibiting glycogenolysis, and finally degrading insulin. In the last step, that's when IDE also gets degraded, thus lowering its total quantity inside the cell. This is completely normal and in fact it's essential, there must be a way to lower insulin once insulin has done its various jobs, but there must also be a way to shut down this mechanism once we lowered insulin back to normal level.

In low-carbers, there's little carbs, so there's little insulin being secreted, so there's little IDE needed to degrade this insulin. But in a sort of contradictory way, there's more IDE in a low-carber than a high-carber, because more IDE is needed to keep insulin lower. Or if we prefer, there's more IDE because there's less insulin to degrade, and the production of IDE is mostly constant independent of external stimuli. As a low-carber eats more carbs, there's more insulin being secreted, there's more IDE being degraded, there's less IDE in the end, blood insulin stays higher than otherwise, insulin can now handle a bolus of 75 grams of glucose, without any significant portion of this bolus lingering in the blood long enough to cause the doc to diagnose insulin resistance, i.e. to diagnose diabetes type 2.

In a way, it must absolutely work this way. Otherwise, a tiny amount of dietary glucose will cause very serious problems. IDE inside the liver would stay higher than otherwise, insulin would get degraded too much, there would not be enough insulin in the blood to deal with excess dietary carbs, we would suffer the symptoms of diabetes type 2 almost instantly at every meal. Or, the pancreas would need to produce more insulin than otherwise to compensate the excess IDE.

Granted, it's just an idea, not a fact, but it's quite plausible that it works exactly as I described above, so in a way it predicts a drop of IDE in the liver when we eat carbs, which should be detectable in the lab by some genuise scientist. And if it's already been detected through experiment and a paper has already been written about it, then that's all it takes to confirm or refute the hypothesis.

For our purpose however, we're talking about dietary fat causing insulin resistance. Dietary fat cannot cause insulin resistance, it does not have the physiological means to do so. However, it does have the means to activate the PPAR-x pathways in the liver, which also happens to be the same way we activate IDE in the liver. Well, if we activate IDE, there's more IDE, we degrade more insulin. And if we degrade more insulin, blood glucose is now allowed to linger higher and longer than otherwise, thus giving the appearance of insulin resistance, i.e. cells do not respond to insulin, therefore blood glucose lingers higher than otherwise. But as I explained, insulin resistance does not exist, so the appearance of insulin resistance is an illusion created by our misunderstanding of the actual mechanisms, which my hypothesis above should help explain more correctly.

This is just how I see it, not necessarily how it actually works.