Which sight is the SOS? Are you calling it by its in game name or its real life counterpart name. Is it a modded sight?

That is the in game armory and inventory name listed by SOS and optic_SOS on
https://community.bistudio.com/wiki/Arma_3_CfgWeapons_Items

The LRPS ( optic_LRPS ) on the same list also presents with the issue

No mods used in the test environment just the Projectile Path Tracing script

Oh, Thats weird. In the dev branch it's called MOS (Marsmen Optical Sights) not SOS.

The crosshair is totally different (no mil dots) and magnification is now "2,5x-5x".

I guess the new sight make up for it. That and the LRPS and the SOS were identical in performance.

Yeah on the dev branch they changed the optic entirely. the MOS is basically a new optic with the SOS's old model and the SOS is no longer there. when the added the Rahim that also added the LRPS that uses the same optic model the SOS had minus the SOS's alternate top sight. So on the dev branch its the LRPS also had the same issue when I tried it.

It's because the new weapon-based initspeed values aren't taken in to account in the zeroing calculations that the game does. The trajectory calculation used to resolve the muzzle elevation when adjusting the zero only uses the initspeed value from the magazine, not the final muzzle velocity of the weapon.

30Rnd_65x39_caseless_mag for the MX and the derivative magazines for the Katiba have:
initSpeed = 800;

The standard-length MX (arifle_MX_F) has:
initSpeed = 800;

The Katiba Carbine (arifle_Katiba_C_F) has:
initSpeed = 820;

These classes follow the zeroing calculations quite well because weapon-initspeed is the same as (or very close to) the 800m/s value used to resolve the zeroing calculation.

The Katiba Carbine POI might show a very small deviation from the POA at longer ranges because it's 20 m/s faster, but for shorter ranges the trajectory wont have deviated enough for you to noticeably miss the target.

As for the other weapons...

Long-barreled Katiba (arifle_Katiba_F) has:
initSpeed = 900;

MXM (arifle_MXM_F) has:
initSpeed = 920;

MXC (arifle_MXC_F) has:
initSpeed = 720;

These weapons have an initSpeed value that deviate greatly from the 800 m/s value used to calculate their trajectory for zeroing, therefore the POI shifts from the POA quite a lot.

You can see that the classes where weapon-initspeed > magazine-initspeed are the ones that shoot high, and the classes where weapon-initspeed < magazine-initspeed are the ones that shoot low.

Another game breaking issue.. upvoted.

The same bugged zeroing behavior can also be observed with the TRG-20 and the (A)RCO optics.
You have to aim significantly lower than expected.
And it matches the same pattern with a largely deviating initSpeed between magazine and weapon:

30Rnd_556x45_Stanag has:
initSpeed = 920;

TRG-20 (arifle_TRG20_F) has:
initSpeed = 840;

And as ArmA is mostly about Infantry, not correctly zeroed optics are a game breaking issue which I'd like to see fixed.
So BI, pretty please...! ;-)