How do we know that criminal behavior does not change the brain, like digging with a shovel creates callouses on the hand?
You've given a good example of how he MAY BE confusing cause and effect.
Still, it's an interesting hypothesis. If it were true we could dispense with the trial, just give the accused a brain scan.
Of course we'd have to neglect all those pesky people, who, like the researcher, have the reduced frontal lobe activity and are not known to have murdered anybody...