Lol dude. If you want to compare car deaths to ww1 the most broad you could be would be to pick a given 4 year period (the length of WW1).
But 1914 to 1918 were particular years, with particular car usership. So are you trying to compare car deaths between 1914 to 1918? I doubt it.
So, you are really talking about needing a shared denominator. Time is convenient. So we look at ww1 as deaths/war/4(years) thus you have relative deaths/year for the given conflict. Now you can take those figures anywhere, and consider car deaths from any year.
So you can then say things like: "car deaths in (year) were on par with ww1 deaths for a year" which is a consistent, logical, and non bait statement. And also has quite a bit of gravity on its own.
If you don't constrain for a shared unit of time, then all bets are off. Old age has killed orders on magnitude more people than cars. Heart attacks while shitting has probably killed more people than cars. It's the timespan that really matters.