WOLF3101,
The bottom pane of a glass tank is designed to float on the plastic rim...bedded in silicone.
The glass tanks I have looked at seem to have the bottom plate laying directly on the plastic rim. The silicon was on the top side attached to the verticle panes. Maybe the high-end tanks have the silicon bead between the bottom and plastic frame. Although I'm not sure how much that bead would absorb any twisting force. A 3/4" wide bead of silicon would compress far less than a 3/4" wide piece of foam.
In other words a single point of a 6 lb rock knocked over by a fish can break a glass panel that a 200lb man can normally stand on.
I agree, the lbs/inch force of a rock hitting the bottom is far greater than a stationary man standing on the bottom (even up on one leg). However, the 6lb rock would weigh considerably less under water (depending on rock density/water displacement) and travel slower through the water as it falls from the top of a rock stack. Impact failures are a different story; nothing to do with an unlevel surface that a tank sits on.
A couple of other failures I looked at had ENTIRE intact panels laying on the ground...this was obviously a bond failure from poor construction.
Actually a failure from a twisting force could also "pop" an intact panel off onto the ground. You would draw the same conclusion that it was a bond failure. The question is, was the bond failure from poor construction, or from a twisting force that exceeded the design specification for the seam, or from both (a weak spot in the seam that would not fail on a perfectly flat stand AND a higher than expected twisting force)?