"Is it? I clearly say that the intact floors are just hanging on the columns; like pictures on a wall." This is the crux of Heiwa's lunacy. He thinks that the floors can hinge, and still support the entire weight of the floor (and maybe even their contents due to friction) and the columns, now unbraced and subject to extreme lateral forces, will not collapse.
Reason why a floor can hinge around a column is that there is a third element between floor/column, i.e. a connection, in this case some bolts and an L-angle.
So if the floor fails due to overload, the load drops away, and the broken pieces of the floor with no load on hit hinge around the connection. Any bracing remains in place.
If connections on one side are broken due to overload, the load drops away and the floor hinges around undamaged connections, where bracing again remains.
In all cases you must analyze where the lose end(s) of the floor end(s) up. Probably on a floor below. The broken floor hinges down around the column connection and the free end contacts a floor below.
Evidently all connections of a floor cannot fail simultaneously due to overload. The floor is connected to four outer walls and to four sides of an inner core structure and there are plenty, say 700, bolted connections. I know NIST suggests that 6 floors suddenly come lose - 4200 bolted connections failed - but that is the usual NIST nonsense.
I have described it all in my famous paper at
http://heiwaco.tripod.com/nist.htm . More to come in the ASCE Journal of Engineering Mechanics soon.