Very thought provoking thread. Thank you for the time you spent on it.
Actually, the calculations sort of highlight the point that it's rarely necessary to physically disable attackers--that they tend to run. The probabilities show that if it were actually necessary to shoot all attackers to the ground, the chances of success with a typical carry pistol are miserable. Given that we know that defenders succeed in multiple-attacker scenarios on a fairly regular basis, we can see that something else is happening a lot of the time and that something is attackers often choosing to stop attacking when the shooting starts. That happens a good percentage of the time.
Focusing too much on the random attack may be a mistake.
Some assailants know their victim and plan an attack on a specific person days or weeks before the crime. Such assailants would be more willing to fight it out seeing how the reward for the crime is perceived to be higher than the random attack. For a random attack, the victim may or may not have anything of value on his or her person, so the assailants aren't as willing to fight it out as the reward for the crime is not guaranteed.