The obvious market and proving ground for a working "smart gun" is law enforcement (gun always on person when on duty, purchase decisions are by public agencies, etc.)
So, when we see a couple of years of deployment of such a technology with major police forces, we can see if they work and are reliable.
I really don't see these things doing well in the marketplace, just based off of who buys guns.
Would probably be fun to hack, though.
Is that because people who now buy guns already believe they are safe when used correctly?
Well, I would argue they are safe when used correctly. The same way a table saw is safe when used correctly, doesn't mean plenty of people aren't hurt by them. I would say it's more political than that. Gun owners don't seem to be the type of people that would be excited to strap a microcomputer onto their firearm to decide when it can be used.