I'm going to play devil's advocate. I think it's possible for the damage to have occurred as reported. Here's why...
The frame on an iPhone 4 is thin. Very thin. Really just a band of metal around the perimeter of the phone. There is an internal rigidty plate, but it is not integral to the actual frame. It would not take much force to bend it.
The front and back glass plates actually contribute most of the device's structural integrity. Assuming that an impact to a cased phone could break one of the glass panels (and I believe that's possible, also), you'd instantly lose almost half of the unit's rigidty. Furthermore, the remaining attached glass shards would act as damping masses. Although that sounds good, it is not from the frame's point of view. It would make the bending forces much worse. Think of the way you'd carry twin buckets on the ends of a bamboo pole like the way they did in olden times. Very bendy.
As far as the cracking of the glass itself, it is quite viable under certain conditions, namely a multiple-impact situation. To crack a device protected in a hard case, you'd either need a single impact of considerable magnitude, OR two impacts in rapid succession. It's difficult to explain in detail, but basically the second impact meets the device as it's on the return waveform from the initial strike. Think of the two buckets on the bamboo pole again, and giving a karate chop right in the center where maximum flex is occurring. This multi-impact scenario is consistent with the clattering-type drop the OP described.
Also, OP described the impact on the tile as "flat". The direction of trauma would be correct for this type of damage, i.e. at a right angle to the breakage plane. A just-on-one-edge face-plant would create a nice bending strain.
I work with vibrational harmonics, and though I don't claim to be the greatest world-class expert on such things, I do have a feel for frequency physics. Just my .02.