Remove AA gain from the Pcal correction method

We normalized the AA transfer function in O3 but there was confusion about this both in the lead up and during O3. I had it correct all along, but thought about leaving in the gain for the Pcal correction in the new pyDARM. After updates to the subway map, I'm content to leave the gain out of this computation because it is the correct thing to do

Note that when I made the original unit test that I specifically left the AA gain in (this is not correct). To verify correct math, check G1501518-v21