Faster? You really mean that? Please explain.
http://en.wikipedia.org/wiki/Buckling
Yes, I mean it. Should I copy the explanation from post #1?
I try it again because R.W.Fassbinder said: If it makes effort then it's always better.
Here it goes:
1) The center part of the core dropped. That part needed about 0.17 seconds to litarally fall at gravitational acceleration.
2) At the same time the perimeter wall (north face) started to bow inwards suggesting an intact floor system that pulled the wall towards the core.
That motion caused an optical illusion seen from the vantage point of "camera 3" because in the 2D video it looks like some bowing downwards.
3) Several frames later the bowed north face gave way and dropped. It reached immediately an acceleration of a little more than 9.81m/s² for a little less than a second.
Imho core-floors-perimeter acted like a spring system. In other words, the perimeter wall was accelerated by the already falling core (and the core was decelerated in the same way).
Of course the center of mass of the entire system didn't fall faster than free fall but the perimeter wall did for a short period of time.
4) the perimeter slowed down to 9.81m/s² and fell further as a unit with the core. That happens exactly in the moment when NIST's "stage 2 - free fall" started.
Hence, since I (and nobody else) have no problem to see that NIST must have measured free fall for "stage 2" the only possible conclusion for the measured deceleration is a "faster than free fall" drop of the north perimeter.
A secondary conclusion is that NIST interpreted the visible movement in the wrong way (besides other failures in their method).