Their cohort studies are across three completely different populations. I would be willing to listen to an argument that for some specific population there’s a firearm ownership level above which there are diminishing safety returns, but I find the idea that this same level would hold across wildly different populations absurd