Touching a touchpad

We've been treated to a number of close races in the Olympics, along with a number of underwater shots of swimmers finishing their races. What I've been surprised at are the number of swimmers who put their hand into a "Halt!" position as they come to the wall. While the added resistance of that position is probably not going to hurt more than 0.01 seconds, the fact that the touch is delayed by about 6 inches of arm movement (your arm+hand is about 6 inches shorter in this position than if the fingers are pointed). That's at least 0.07 seconds (for 100m in 50sec). For instance, see the finish of the men's 4x100 Free Relay (in this case, the position of the hand loses only 2 or 3 inches), or more dramatically, Hoff's finish in the W400Free where the touch technique may have cost her the gold. Is there some reason that swimmers are finishing this way? Why not drive your fingertips into the touchpad? Related to this, can anyone point me at documentation that shows that touchpads accurately record touches to at least 0.005sec, no matter where on a touchpad it is touched, by a sharp jab or a soft, flat hand, and that any two touchpads have the same latency. I've seen what appeared to be discrepancies between recorded time differences and what video shows as the difference in two touches. I did find it interesting that at the World Masters meet at Stanford, relay starts that had a reaction time of -0.04 were considered not a false start, but -0.05 were a false start. If these were perfectly accurate, why isn't -0.01 a false start? Any documentation on this?