yeah, that should be possible to do, I think. If we line two choruses up, and sync the modulation to each other, only invert one of them, that'd be enough, I reckon? Perhaps the modulation depth in the second one needs to be deeper. My rationale is, if you pitch something down by 10%, to then pitch it back up needs an 11.1111% increase.
Try posting your patch so far here, then we can hack at it.
I'm curious, what's your intended goal here? To hear the artifacts created in the process? If any I'd imagine it's a downsampling of sorts?