Still experimental, but in the case the Warlock uses Scourge very shortly after being struck by redemption, he will teleport back with reverted knockback (so that the situation would be identical with a normal Scourge-Scourge situation).
Seriously, what's the reason for this? If inquisitor scourged first there's no reason for the reverse effect to be cancelled out. Not to mention that this effect is ridiculously abnormal for a physics-based game like Warlock. It'll only serve to confuse.
And what if a double-scourge occurs with the warlock near lava and inquisitor near middle? The inquisitor scourges first, warlock gets tped toward the middle (which was very likely his plan), then scourges, hoping to push inquisitor to the middle, but finds out he's back where he started?
Or what, am I not understanding this correctly? It sounds inane.
What I am trying to fix is that a situation can occur where both Warlocks end up having the same knockback direction, which is both silly of nature and inconsistent gameplay-wise as a millisecond will completely change the outcome of a double scourge event.
The scenario that you described already includes an error of thought.
Assuming it was the Warlock's plan to get pushed to that side, then how would he ever be able to control whether the inquisitor should follow him (same knockback direction) or be pushed away?
With the fix, you will know that the double Scourge will always have one outcome, whereas now, it's 50/50 if you end up having the same knockback direction or not.
As for your scenario, if the Warlock plans to get to the other side of the Inquisitor, he simply doesn't counter Scourge. You have about one second to figure out whether to stop the Scourge or not. For the current knockback direction, you have about 30ms to control the event.