It's well documented that in 1967 or so the AGC code was bloated (amongst other problems) and looked like it was not going to be ready in time for the landings, so much so that NASA sent in Bill Tindall to MIT to kick heads. Could they perhaps have given the under-pressure programmers some breathing space - a contingency - by carrying another set of ropes with the excess (return mission) code on them, whilst still working on the all-in-one set? That is, fly to the moon with everything required up to P65 etc then once on the surface, exchange the rope modules for the return software and throw the first rope set out onto the surface to save weight. Power cycling the AGC in flight was possible and even done later on Apollo 13 and surely they would have done this in simulations. And they could presumably have left the IMU running and aligned, as sufficient power was available? Steve