Conceptually, it’s a true thing to say.
And it’s the point I tried to make as well in the twitter thread when I referenced saturation rates. And it’s also the point I made in here.
If we were somehow able to strip all the multivariate cofounders away, which is nigh impossible, and then plot a true best fit line, then it would start at the origin, run up steeply, and then flatten out tremendously. After some ratio of gun owners (could be as low as 10%? 5%?) it would be dead flat, because scarcity would no longer matter at all.
There’s no good way to draw this line.