@Jamie,
I understand. For the reasons I explain above, I think a case could be made for it for the reasons I give. If one really wanted to get fancy, you could run regressions across countries, tracking laws, mental health, weather, population density, and a host of other variables, and then try to figure out the impact on suicide of both guns (on civilians) and war (on soldiers), and then count as many of the suicides seemed actually to be due to guns (or, respectively, war).
But we're arguing on an internet forum, not getting econ Ph.D.s. For our level of rigor, either your approach or mine is likely fine.