I'm also not sure why you think I'm "misunderstanding" the BMI concept, because it is painfully simple--rather, I think it is you who is misunderstanding what I'm saying.
I'm talking about BODY COMPOSITION for a particular BMI and the contribution of LEAN MUSCLE MASS to that body composition.
One more time:
Because women have both smaller bones and often lower bone density than men, for a given BMI, a woman will have less of a contribution to total body mass made by bone mass than will a man; this means more of a contribution from soft tissues, which means more soft tissue bulk.
Militating against that is the fact that because women have a higher % body fat than men, for a given BMI, a woman will have less of a contribution to total body mass made by soft tissue mass than will a man; this means less of a contribution from soft tissues, which means less soft tissue and bulk.
IF those 2 factors equaled out, all BMI-related things would be equal among men and women. But DO they equal out? I don't know, but I don't think so. It is undeniable that trained sprinter women are much leaner than the general female population. It is also undeniable that regular exercise generally increases bone density among active, pre-menopausal women.
In my estimation, the first mechanism dominates, significantly. These women are lean. For a given BMI, they have more lean muscle than do male distance runners. Even more than that, that lean muscle mass is distributed to the quads, butt, and hips, exactly where it is advantageous for sprinting--totally unlike distance men.
And further yet, only some of the best sprinter women have BMI's that reach down into the elite male distance runner territory.
And most of all, it doesn't matter, because it is a matter of neurology and training, which brings me to the last point, which is that the more highly trained an athlete is for endurance running, the relatively worse that individual will be at sprinting. PERIOD. END OF DEBATE. NO DISCUSSION.
There is no doubt that the best-of-the-best distance men are highly trained, therefore they are as bad at sprinting as they could possibly be, from a running training perspective. Massive distance training does NOTHING in service of sprinting--it does not train any form at all, it does not train power, it does not train acceleration, it does not train transition, it does not train maximal speed, it does not train the right energy systems, it does not train absolute-speed-endurance, it does not train proper ROM, it does not train maximal turnover, etc. It trains NOTHING of direct benefit to sprinting.
Once that massive aerobic base is built, coaches can try to turn their attention to using whatever short-term reserves are left in the muscles after that 9,600 or 4,600m, to extract maximum possible performance in the last 100m, or last 400m--however, compared to sprinting and sprint training, that is paltry, and only trivially develops capacity for a fresh sprint.
This is why many argue that these guys wouldn't break 50 for 400m when fresh, even though they go 53 or 54, or whatever, at the end of a distance effort--and I'm starting to believe them. Acceleration and leg speed are still important in a 400m, and distance athletes don't have either.
Here's how I think about it: OK so someone goes 53 at the end of an aerobic effort, over the last 400m with a running start. Fresh, say with whatever running start distance it took to get up to top speed, they could then do what? 47? 48? 5 or 6 seconds faster? Then tack on a couple of seconds to account for the acceleration that would be needed if run from zero velocity rather than from a running start, and you are back at, or over, 50.
Does one need to be able to run sub-12 FAT from the blocks in order to go, say, 50 in the 400, FAT from the blocks? No. Consider these possible race profile for one who could maxx out at 12.5:
13-11.5-12.5-13 for 50 seconds
or
13.5-12-12-13 for 50.5 seconds
etc.
This for someone who could go 12.5 FAT 100m from the blocks, or around 11 for a flying 100m.