This requires a correction of near 30 MOA per barrel. That's half a degree. The barrels shoot to POI near a full degree of arc apart.
Were we to pivot the barrels, depending on the pivot point, someplace between .1 and .25" of movement per side would be needed.
This is not a minor assembly error.
What we don't know is how the barrel set is assembled, by cold or hot solder, brazing, or what combination where. Only the factory could tell us.
The whole fiasco leads me to the conclusion that the barrels were assembled on a jig, scaled down from a larger gauge gun without any sort of testing whatever. It's highly possible that each and every copy of this model has the exact same issue.
