Actually that article seems to be addressing a completely different (and indeed wrongheaded) argument that states that it's fatalities-per-hour that matter, not

Actually that article seems to be addressing a completely different (and indeed wrongheaded) argument that states that it's fatalities-per-hour that matter, not fatalities-per-mile. It doesn't address the real issues here, which are that: a) The assumption that the proportion of miles walked drunk is the same as the proportion of miles driven drunk is completely unjustified, b) They don't distinguish between different levels of "drunk" Whether these arguments are ignored through ignorance or deliberate disingenuousness, though, I'm not sure.