Yes, it is obvious that NIST in their answer to the energy question neglected the all important question of energy considerations related to the collapse, and that the airplane collision is only part of the energy equation. I read with interest the paper by Hoffman on the energy required to sustain the cloud of dust following each of the towers' collapse. But it seems to me that the cloud could be a result of displacement rather than of thermal expansion. Furthermore, while one generally assumes an inelastic collision of the building's components with the ground, this assumption may not be entirely correct, and there may be an elastic component to the collisions that could drive the energy of the dust cloud.

What I believe would be of greater interest is an analysis of the potential energy vs. the energy required to sever the steel beams in the building. We know from the rubble pile that the most of the steel had been severed into small pieces. Even a simple model could reveal a lot about the energy required to sever all the steel beams in the building. There are many ways to sever a beam, but a gravitational collapse narrows the ways in which beams can be severed, and this affects the energy requirements for each break..

I recall a paper that concluded based on seismological evidence that the collapse of the towers was slower than freefall. The difference between the kinetic energy generated from freefall and the energy due to the slower descent represents energy available to break the building apart. This figure could be compared with energy required to sever the steel beams, to see if it was adequate to do the job. Since I have not done the calculations myself, I would rather not pronounce a verdict on this question. 4x10^11 joules x 0.1 (estimating the excess energy not translated into kinetic energy) /2,500 ruptures of the core support beams gives one an idea. Is that enough energy to break a steel beam?

WTC 1 collapse proceeds at free-fall speed, through the most obstructed path of travel - the highly redundant core: