This is a continuation of
https://l2jmobius.org/forum/index.php?topic=14262.0 but with a different (separate) issue
I found out that every debuff cubic will go through at least 2 landing checks:
1. The outlined aboute calcCubicSkillSuccess method (if it fails, the players gets "your attack failed" system message)
2. The second one is in calcEffectSuccess method which will use the player's m.atk for the calculation (if it fails, the player gets "your Hex skill has failed" or something like that)
Even if (2) is necessary, I would probably assume it using the cubic's matk rating. A simple test with using a magic weapon vs a bare hand shows a significant change in the landing rate.
I used printing to debug the exact values, and here is a proof that it uses the summoner's/SK matk rate (an example with no weapon equipped):
[30/07 12:39:16] >> applying Decrease Atk.Speed
[30/07 12:39:16] target level = 57
[30/07 12:39:16] matk = 9.23750784
[30/07 12:39:16] mdef = 96.797
[30/07 12:39:16] matk vs mdef = 0.3453887579294713
[30/07 12:39:16] baseMod = 69.0
[30/07 12:39:16] elementMod = 1.0
[30/07 12:39:16] traitMod = 1.0
[30/07 12:39:16] matkMod = 0.3453887579294713
[30/07 12:39:16] debuffMod = 1.01
[30/07 12:39:16] rate = 24.070142540104857
I don't know about later chronicles, but Interlude and C4 used matk-independent formula. And the phantom cubic really did apply around 80% of time when it's a mob. Here, it's ~0% chance (because the first method is broken, see the link about), plus even if we fix it, there is a very low chance of applying it due to the relatively low matk of SK (summoners tend to have a mediocre weapon too).
Is this second landing rate check is required for already passed cubic skills?
Or if it is, should it perhaps use some other matk score other than the owner's one?
I might be wrong, but the cubic power could be the intended matk for it:
<table name="#cubicPower">282 434 644 919 1141 1394 1673 1975</table>
(from the Summon Phantom Cubic XML data)