I played a bit, because I was curious if real ballistics data will work in RO without corrections.
You remember the data. 105m/s projectile, flying at a range of 135m at 3.5deg superelevation. I estimated the speed at 135m 98-100m/s.
So I took my panzerfaust, set it to 105m/s, set a sight to 2.0deg and 3.5deg elevations (for 80m and 135m) and tried few ballistic coefficients. Those in range of 0.3-0.35 works good.
And with ballistic coefficient of 0.3 I got terminal speed of 98m/s at 135m, and the real-life 3.5 degree sight setting... worked with a pin-point accuracy. It hits a house window at 135m - without any sight tweaking, just taking real world value. The max ballistic range (measuread at Arad which has a high ceiling) when fired at 45deg is about 800m. Either the RO ballistic physics are not perfect, or more likely my crude ballistic spreadsheet is not perfect for such projectiles (I estimated 900m).
But up to 200m everything works ideally and even sight settings from real life works ideally (at 80m seting it was hitting tanks too, though I didn't oberve how precisely).
This is advantage of modeling things as they work in real life, instead of creating some artificial rules (fly ideally straight for some time, then begin to fall like a basketball) for them. You set other real life parameters, and it still works as it should. And it's even not hard to do. IMO it's easier and faster to model reality in most cases, than create artificial game models which have to be tuned and tweaked then... and again for each change of parameters...
If there are ballistic trajectories handled by the game, build in (physics projectile) - why not use them ?