Failures that happened further away from the window would have produced longer delays. But in ALL cases, some part of the upper portion of the towers HAD to have begun to collapse BEFORE the smoke began to pour out of the window.
My eyes confirm this when looking at your video. (
http://img690.imageshack.us/img690/9109/femrnew.gif) With a margin that doesn't require video analysis.
The ball of fire clearly (to me) begins to descend BEFORE, not after, the smoke begins to get pushed out of the windows. My reference is your red arrow moving across the screen. The point at which the smoke starts pouring out has the arrow at a significantly later point in its motion than when the fire begins to descend.
This negates the comment that Major Tom made that "the smoke begins emerging before any part of the periphery of the building begins to descend.
You are free to offer your speculation as to how what Major Tom said could be correct: that the smoke began pouring out before any downward motion of the periphery of the building. But your own gif video seems to dispute that.