The profile shift (rack shift) affects the Lewis factor Y for bending stresses or the bending J factor in AGMA gear strength formulas. For the Lewis factor I have seen a graph with Lewis factor as a function of the gear profile shift in a Russian translated book on machine design from Dobrovolsky. I didn't see something equivalent for the bending J factor for AGMA. The J factor is usually calculated by drawing the tooth geometry. Gear design software usually do it automatically.
Normally, the profile modification has a very big influence in the gear strength. To have a 0.5 deplacement normally means to have approximately doble strngth than in a normal tooth because you are going to have a bigger base on the gear teeth but, (in gear modifications the most important word is but) you are going to modify too the head thickness, and possibly it is not going to be enough to resist the torque. You must see everything while making profile modificatons.
As it is said before, you are not going to have only problems with fatigue force. You must see if the teeth are resistant to Hertz pressure (normally, pitting problems appears before fatigue problems). To have more pitting endurance, you must controle the specific slidding value for pinion and wheel but lot of things have a big influence on endurance.
You must calculate the complete gear train with its external influences (lubrication, forces,tolerances ...) To study only a part is like don't have any study.
