I simply don’t buy this at all.
First, in the factory geometry the lower control arms are not parallel to the ground. The difference in height between the LCA pivot and the lower ball joint is 1-7/8”, for the LCA to be parallel that number would be 0. While it’s hard to tell from just a picture, it doesn’t appear to me that the OP’s car is higher than factory. With the drop spindles that could be changing the lower control arm angle up from stock, as he doesn’t look 2” lowered, but I’ve run drop spindles without that issue.
Second, the ride height change he’s describing is 2”. Even if there was some validity to that scrub radius effect as you’re describing it, that change would be no where near 2”. The toe change required to do that would be massive. More likely the change in toe is following the change in ride height (not the other way around like you're describing), as the car lowers the LCA's get closer to parallel to the ground, making them effectively longer and because the tie rods stay the same length the rear steer set up causes the wheels toe out.
And finally, I run the same QA1 LCA’s, high offset wheels, more negative camber and more positive caster than the OP does without anything like the issues you describe.