Reply #17: You may be asking a decent question... but you are refusing flatly to listen to the answer.

[Realityhack]

There is no 'before the collision' to look at. It isn't like a beam from the death star erased one floor from reality and the top of the building was in free fall for one story.
It may be *close* to that. You may be able to simplify to that *for certain calculations* but that was not what actually happened.
So you can't just say 'well before the collision x happened'. The upper and lower sections were in contact at various points throughout the collapse.
Furthermore as things progressed the upper section began to rotate. That alone would drastically change the loads on the various structural members creating any number of different breaking points etc.

You had an idea, you posted it here, but you are refusing to listen to the answer.
The idea of calculating what those trajectories should be is massively more complex than you think it is. People with a better idea of what the complexities involved are are telling you this.

As jberryhill pointed out, you are welcome to prove this statement incorrect. You can do the math and publish the paper and show it is possible.

Short of that you have no argument.