Just a rough guess, numbers from the back of an envelope.
Current US homicide rate is 4.3/100,000 implies 15,300 murders with population of 340 million - let’s pretend one murder per murder. Let’s guess average age of murderer is about 25 when they commit the act, and lives to 75 (since death row is a long way off, most homicides are not capital punishment. So 765,000 killers in the USA who have not died yet. Actually, the homicide rate is much lower than several years ago, but this compensates for multiple homicides, multiple perps, early deaths…
About 33,000 automobile deaths last year in the USA. Let’s pull numbers out of our butt and guess say, 1/3 of cases the killer responsible survived (since many would include the driver died, and took multiple people with him) So 11,000 killers a year, and let’s guess an average age of 35 at time of act, and life expectancy of 75 again, so 40x11,000=440,000.
What about other causes? Lets say any other causes of death are much less; drowning - typically the victim’s fault, rather than a “killer” responsible. (I suppose some idiot taking an overloaded boatload out on the lake would qualify.) Accidents like chopping a tree down on someone, or such.
So we have 765,000+440,000=1,205,000
Let’s pull a number out of our butts again, and say this is much smaller… lets just round this to 1,300,000 to cover all eventualities. From 340,000,000 people, that’s 0.3%
I have no idea how to account for military “killers”, except guess that say, 5% of active military would have killed someone in war; 170,000 at peak in Iraq, 100,000 peak in Afghanistan; over 500,000 in Vietnam. Allowing for turnover, maybe we’ll say 1,000,000 soldiers. Let’s say 5% of those actual engaged in combat where they killed someone. (Is that too high? Too low?) 50,000 armed forces killers - hardly budges the total versus 1,300,000.