I doubt many doubt that pilot psychology affects the likelihood of unexpected death while flying.
What I am sure would suprise nearly everybody is that GA equipment modernisation has not improved the fatality rate.
It has definitely improved the airline crash rate - though together with this we have much more sophisticated training compared with the goode olde days of "hire all the ex Spitfire pilots".
So what does that leave, as factors affecting fatalities?
It leaves the pilot and his training!
The next level of analysis would try to separate the effect of pilot's attitude from his training.
Neither has changed since GA was invented ~ 100 years ago. The lack of PPL candidate psychological screening has not changed. The training has also not changed (in any significant manner) - it is still treated like training for a hobby e.g. how to grow cabbages... something you can drop out of when you get bored (which is exactly what the vast majority of people do, pretty quickly).
So I am not sure where to go from here. Psychological screening is obviously not going to happen (nor should it). The equipment is not going to magically improve (the world is owned by Garmin). That leaves only the training, and I cannot see anything ever changing on that front - too many vested (business) interests.
We will see better capabilities e.g. precision GPS approaches to places previously without any approach (maybe not in the UK with its money-driven privatised ATC, but elsewhere). But these are prob99 not going to reduce fatalities because whereas currently almost nobody flies DIY approaches, everybody with the kit will be flying LPV approaches. It's like the SE v. ME argument: twins are no safer because people fly riskier mission profiles. This may be true with avionics also; a G1000 and a parachute is bound to reduce the perceived risk (wrongly with the G1000 but rightly with the chute).