>>IMO, if it was the clips that failed because of buckling, there's no way the floors would be falling straight down. The side which failed first would go down first. We'd either see them piled down on one side, or stuck diagonally around the central core. The floors falling so neatly down is just not something you'd expect to be caused by random malfunctions of whatever harness was holding them in place, not even once. Three times is a tad too implausible.
>
>How far above the ground should I hold the book? To be to scale I'd guess less than a foot.
>Why would it need to be perfectly parallel - it ends up flat on the floor and it weighs the same?
>But it could be argued that having an outer edge impact on the floor below first would result in greater stress on those clip points.

Desk level would do, so you could observe, in freeze frame later, how did it go down. My bet is that the number of times it would fall horizontally would be close to zero.

>Also, who said they were 'random malfunctions' ? For the two main towers, as I read it, it was an inherent (but unforeseen) weakness in the design.

OK, so they were deliberate malfunctions, whole floors' clips failing simultaneously, not one ahead of the others.