A conversation at lunch reminded me of a paper/abstract I read tied to fat consumption and total insulin (area under the curve) - my recollection was that the example used something nasty like a baked potato with and without a huge glob of butter.
Without butter, you hit a huge spike, as expected. With the butter, things aren't great either, but the spike is blunted a bit due to the fat, presumably slowing absorption or whatever.
But the punchline was tied to the total amount of insulin used was 3x or something ridiculous in the butter (fat) case. So with just the potato, you spike higher and that high BG is bad, but with the other case, the amount of insulin released into the system was much higher despite having no spike.
The studies and reports tied to the difference between high insulin and high BG are many, but I can't find the reference to the total insulin being "worse" with the non-spike case.