Remember that "+11.75% hitstun" doesn't actually mean "+11.75%". The game's default is 0.4 and "+11.75%" is 0.5175 . That means that "+11.75%" is actually 29.375% more than the game's default hitstun (it is 0.4 * 1.29375). This discrepancy is the result of an inaccurate naming convention. When we say "+11.75%", we actually mean that we are adding 0.1175 to the game's default value of 0.4 .
We don't want to keep the ratio of fallspeed to hitstun because the ideal level of hitstun is inversely proportion to fallspeed, so we need instead to keep the ratio of 1/fallspeed to hitstun identical. So, if you made fallspeed = 1.35, by your logic hitstun should be multiplied by (1/1.35), because that would keep the ratio of 1/fallspeed to hitstun.
1.29375 * (1/135) = 0.95833
0.95833 * 04 = 0.383333
So 0.383333 is the number you should, by your logic, make the hitstun at 1.35x fallspeed.
Remember that the hitstun should be inversely proportional to the fallspeed, so
Current ratio of 1/fallspeed to hitstun is (1/1) : 1.29375 = 0.7729 : 1
New ratio of 1/fallspeed to hitstun is (1/1.35) : 0.95833 = 0.7729 : 1
Your logic, however, is flawed because fallspeed and hitstun are not directly linked in a 1:1 level of importance/weight like that and keeping the ratio will not yield the results you desire: we will simply have to test it a whole lot more. May I suggest we find the ideal hitstun at two different fallspeeds and then plot a graph by drawing a line between the two in order to find the ideal value of one for a specific value of the other.
Sorry if this was confusing
.
Also: vis-a-vis the fallspeed code having effect while in knockback: maybe the gravity used while the character is in hitstun is the falling gravity. That would explain nerfed vertical KOs (which, I might add, I think are a good thing).