So, to summarize, the first suggestion was pancaking floors like you would expect in any progressive collapse. I then pointed out the AE have shown the explosions do not progress down regularly. Mick prefers to imagine whole floors collapsing, but in small rooms. Others diverged and went down the pieces of masonry route. To escape the inevitable conclusion that pieces of falling masonry won't generate enough of an increase in air pressure to blow out windows, we then have solution 1, that says the masonry hit the floor below and that was what generated a rush of air, and we have solution 2, that says no actually, it was falling streams of fluid-like rubble. Why not include the hypothesis that explosions and molten metal might indicate the involvement of incendiaries/explosives? It costs nothing to admit the hypothesis, but insisting it be rejected before it is considered comes across as biased.