Sorry, but TFA has it all wrong. This is about values, not ratios. The question at hand is whether you think these deaths are acceptable given our values as a society.
To put it in nerd-speak, the intolerance for any given death is a function V with multiple inputs. (I find this extremely distasteful but bear with me.) For cars and smoking, the output of that function scales quite slowly; we're mostly OK with those because we like cars and if you smoke it's your funeral. Note that we already spend a lot of money on automotive safety, so arguably we still want to bring those numbers down. For instance, look at recent safety-oriented recalls from various manufacturers. Likewise, we spend money to reduce teen smoking, et al. So even then those don't really wash as an example of hypocrisy around the value of human life.
For the murder of multiple people via gratuitously overpowered firearms, you might argue V scales linearly, perhaps even quadratically. Maybe V takes a time-delta which scales V even faster with smaller deltas; Aurora is still fresh in many people's minds, as are numerous other incidents.
For multiple children under the age of 8 gunned down by a madman with a high-capacity assault rifle, V might scale factorially. We as a society try to place a very high value on the lives and well-being of children.
In other words, you're going to have to do a hell of a lot better than simply comparing inputs to V, and a small input to V for some category of deaths is not in itself a justification to ignore it. The author has committed the fallacy of looking at morality and society primarily or even solely through the lens of statistics.
The author's laziness is also ironic, all things considered. The UK and Australia are other western industrialized nations quite similar to ours, in a great many respects. They've experienced some success in reducing gun violence since the '90s, when they banned private handgun ownership. The irony is that the author's blindness here is a far better example of magical thinking about how regulation might or might not work in the US.