Driving to work this week I heard another radio interview, done a little later than the first I heard. The updated information I got from it is that they could not leave the scene. The car was boxed in by panicked shoppers. He could not have just driven off to safety. They were stuck. In all probability that realization may be what made him act.
Ahhhhh! More information. Good!
If that is indeed true -- if they really couldn't leave by simply driving away -- that narrows and alters the set of other "reasonable" responses.
Would getting out of the car and fleeing on foot be a good idea? Quite possibly. Out the lee side doors, stay low, move quickly and steadily to better cover, hopefully perpendicularly to the lines of battle.
Or, quite possibly not. If they were close enough to be screaming at the goons then getting out might be leaving even the meager cover provided by the car.
Would staying put be the right choice? Wow, that's a tough sell, because most of us know what a car door WON'T do to stop a bullet. Of course, depending on angle and proximity, it may indeed have been the right choice, if the general direction of fire was parallel to your location, it might be substantively safer to stay put, get low, and hope for the car's sheet metal to slow or deflect any glancing hits or ricochets for the next two or three seconds that most gunfights actually last. Certainly IF that were the case, getting out of the car and calling attention to yourself would probably be the wrong choice, as drawing fire toward your car and family would make things much worse.
And, the truth is, being trapped in extremely close proximity to something like this might be one of the very definitions of being "engaged" in the fight without much choice in the matter.
As always, drawing a gun is what one does when there are no other viable choices left. If this guy was trapped with his family in the midst of a gunfight, he may have had no better choices.
That's not how he and his wife spun the situation in interviews, but the gulfs between "what happened," and "what someone actually said in an interview," and "what got written in the article," are generally pretty wide.