I think what Heiwa is overlooking (and Apollo20 too, if he's serious), is that impact is a function of velocity and time together.
Heiwa seems to assume that if the upper block is not "solid", it must be macroscopically fluid, and thus would behave the same as water in stream, and break around the object.
At lower flow rates, you have a small mass of water impacting over a relatively long time, which allows the water to flow around the object.
At higher velocities (like, say, the deluge from a water bomber), the mass of water is huge, and it impacts over a very short time. It packs a helluva punch.
The same thing would happen with a large mass of rubble (rubble for Heiwa's fluid upper block). A lot of rubble (whatever consistency you want) impacting over a short time (and in a small area, since most of the mass is vertical, not spread out horizontally), will impact with a force not appreciably less than if it were rigid.
F = m(dv/dt)
F = impact force
m = mass
dv = change in velocity
dt = change in time
so, if m = 33,00 tons ~ 30x106 kg
dv = vinitial - vfinal, which I shall arbitrarily (since I don't know the speed at which the upper block fell) set to a conservative to dv = 5 m/s
dt = time of impact, which I shall arbitrarily set to a generous 1 second
F = 30x106(5/1) = 150,000,000 N
Put another way, that's the same as getting hit with a 1 kg cannon ball (over an impact period of 1 second) traveling at 150,000,000 m/s (half the speed of light!)
It is not an insignificant amount of force, and you can't just wave your hands and say "it never hits the towers", because even assuming the "fluid" upper block is deflected by the lower block, you have to account for the force applied to the lower section as it deflects all that mass over the duration of the impact.