
This assumes that in stressful active shooter situations, bystanders are going to successfully take down the shooter. It’s highly possible that in many situations more gunfire could injure more people. It also ignores that more people with guns means a higher likelihood of these attacks happening in the first place.
There’s another study from a similar time period which says the rate of guns is more like 5% and half of those were security guards or off duty police. Usually attackers were stopped without guns, granted this dataset includes a small number of knife and vehicle attacks but it’s mostly shootings