You don’t need to change the overmatch system. Just use the penetration in WWII ballistics to go along with the slope modifiers already implemented in the game.
Before the calculators were a thing ,Gaijin intentionally nerfed the T33s flat pen without doing anything about the now terrible sloped pen. The system they use to calculate sloped pen needs to have it’s original data inputed or you’ll get the wrong results.
Besides, Russian APHE still has their uber special slope normalization.
How does AP penning more from 0-40 degrees matter when APC is already over performing across the board? Proper APCR performance would out perform AP from 0-40 degrees anyways.
There is was no “original” data. It was all just calculted data but from WW2 Ballistics: Armor and Gunnery.
A slope modifier only works when you consider “theoretical” penetration values.
At high velocity, AP rounds will deform and have reduced penetration against vertical and near vertical armor.
The whole reason why APC was used, apart from the obvious benefit against face hardened armor.
APC isn’t overperforming across the board. A few rounds penetrating more than they should due to Gaijin only looking at the weight of the shell as a whole, doesn’t compare to practically all AP rounds that would penetrate more 0-40° armor.
APCR and APC fulfill practically the same purpose, to penetrate thick armor plates that you wouldn’t be able to penetrate with AP.
Atm APC is superior to AP because it has more 0° pen while nearly the same slope performance.
That missing 0° should be working extra for +45°. Making AP better against sloped armor.
Hence why a simple adjustment to slope and overmatching modifier would be the simplest solution for proper working AP.
I never liked the whole idea of slope multipliers in the first place. If you are gonna calculate the 0° penetration of all shells, and then estimate their penetration at X° by applying slope multipliers to that estimate, you might as well estimate the sloped penetration directly from shells mass, caliber and speed like you did for 0° values, without introducing another source of possible error with slope multipliers.