Long before you try bending the frame you need to better understand what is not straight/square/right. An assumption/base line is needed somewhere and a true and dished wheel with a straight axle is the usual. It seems that the OP is saying this is what he has done.
As randyjawa said flipping the wheel around will help confirm the wheel's not the problem. Next up is determining if the main and rear triangles are on plane/centered WRT each other. Most use the "string test" to check this out. If the wheel sits a bit off at either the seat or chain stays one should be able to sight along the rim's sides and along the main triangle's tubes to see if that offness is only the stays being non symmetrical or if the dropouts hold the wheel a bit cocked.
If the wheel looks off at the stays but is in line with the main triangle then the stays are somehow the issue. It is possible for the seat stays to have a bow to then that placed the brake mounting hole off center. This and the rear triangle being not centered (the drop outs not the same distance from the main frame's center plane) are the only aspects that any bending of the frame will effect.
If the wheel is off at the stays and also is off WRT the main triangle then the drop outs are not located equally in the remaining 2 planes they need to be. A seat and/or chain stay is longer then the other. No amount of bending (other then bending the long stay so much it's effective length shortens
) will fix this. being able to reposition the axle end within the drop out is the cure. Filing the long side's drop out slot allows this cocking the wheel back to being straight.
Of course before and after doing any frame aligning the drop out parallelism needs to be confirmed. The Campy "H" and Park FFG-2 tools are for this. Andy