I’ve seen it suggested on the MakerBot Operator’s group (I forget who…) that Z axis rod wobble can be introduced where the bearings fit on the Z axis rods.  This is an interesting possible source of error… and a solution!

It would seem that if the placement of the bearing on the rod can introduce warp, it might be able to slightly counteract warp as well.  I’ve gotten my Z axis wobble mostly under control.