It wasn't all totally black-box and statistical. There is a lot of analog computing with carefully calibrated theory and decades of data behind the artillery used in WWII. Lord Kelvin & early wave / Fourier analysis applications. Really interesting stuff.

But yes, you are right. I'm pretty sure the early way of doing "stats simulations" was just basically trying a bunch of stuff and recording the outcome. The two approaches are important, I think. There is a lot to be said for "American ingenuity" in WWII where the Axis powers thought Americans were absurd and often had backwards solutions to things, but hey unorthodox works