Ah alright, I see what's going on. So when I used the equations for torque in the OP, I was calculating B, and as such point cg would have been negative, but as we are now calculating point A, cg becomes positive, so in effect it is FB - cg and not + cg.
This changes the distance to 0.8776m...