there are a few separate convos here but I'll stick with just my question...
It's not as simple as saying the thrust is at the nozzle; thrust is against the inside forward chamber of the motor. If thrust was just at the nozzle, it would be pushed up into the chamber. The nozzle pushes on the chamber pressure, but does not exceed it; the chamber pressure pushes on the interior forward wall of the chamber. When a balloon ejects air, the thrust is not solely on the rim of the "nozzle", and is not transferred through the wall of the balloon (which would buckle the balloon); the net unbalanced force is on the inside of the forward wall of the balloon as the mass escapes the "nozzle". So too with a rocket motor.
However, that debate isn't relevant to the question. I'm granting (granted, granting, continuing to grant) that we can treat the entire motor as a solid unit, so it doesn't matter where the thrust is generated on the motor itself. Imagine the entire motor case is being accelerated upwards, without regard to the source of the acceleration. A magical moving cylinder. I am inquiring about the transfer of thrust from the magical moving cylinder to the airframe. In one scenario, the motor pushes on a forward bulkhead/retainer; in the other scenario it pushes on the bottom end of the airframe. Consider the thought experiment in which the threaded rod is centered in the motor, but is dramatically offset to one side on the bulkhead. Imagine it all the way at the edge, so the threaded rod is on a stupid diagonal. My picture is that the motor pushing on one side of the bulkhead would lead to a torque around the center of mass. In practice, the threaded rod is nicely lined up, so this wouldn't be a problem. But my thought experiment is intended to illustrate what I believe is a difference between the two scenarios, e.g. thrusting on the forward bulkhead vs thrusting on the aft airframe edge.
Now, both the motor and the bulkhead are constrained by the airframe to axial alignment, so maybe even a stupidly diagnoal rod can't exert a torque; perhaps that is the answer to my thought experiment. But maybe it would exert a torque, if the rod were stupidly misaligned. I can't sort that part out in my head.
Obviously it works, because people do it, as reported multiple times in this thread. Not debating that, just trying to understand the details of the load path.
Also, I guess it requires a significantly stronger forward bulkhead/retainer when it is to be used as a thrust path to handle 100's-1000's of N's, than if it only needed to handle the force of retaining the motor from dropping out. Do people take extra efforts to strengthen the forward bulkhead when it's used for the thrust path?