Pedal power meters do need to know orientation, which is why they’re supposed to be calibrated at a particular point in the circle (typically, either at 12 o’clock or 6 o’clock). Shoes or shoe attachments can slightly change the “pivot point” so you may not get zero force vectors in the intended direction–but unless the pedal attachments add a bunch of stack it probably isn’t very large. Still, for proper calibration, you should remove them.
As an aside, early models of the SRM had only 2 strain gages and you could get pretty different torque readings around the crank circle.