The statistics specifically address the situation where the victim "RESISTS with a firearm".You would have to be able to show statistics on how many times a victim presented a firearm and how many of those events ended in injury of the victim.
In the general case that offers a better chance for remaining uninjured than compliance.
That is incorrect.Those numbers would then have to greatly exceed the percentage of people that just do not get injured without resisting.
The only thing that the number of cases affects is the margin of error in the resulting conclusions.
In other words, if there were a million cases where a victim complied and with victims remaining uninjured 75% of the time and there were 2000 cases where the victim resisted with a firearm with 83% of the victims remaining uninjured you can still say that the odds favor resistance with a firearm.
The difference is that the compliance figure (75%) has a margin of error that is virtually nil while the the resistance with a firearm figure (83%) may have a margin of error that is a percentage point or two.