hmm #2
The high pressure in an engine or jet would tend to push the back end off: nothing to do with drag force on the nozzle surface, just pressure acting on the rear surface. Nozzle turns pressure into flow speed. Blow up a balloon, turn it around so it faces away from you and let go of the untied end. Would it fly away from you owing to drag over the aperture or into your face because of the thrust ? Put a weak joint between the nozzle of a rocket motor and the case and see what happens (owing to the pressure inside the case) - from half a mile away that is. And the adiabatic expansion or compression on flow through a cone would cause no discernable heating or cooling (convergent at subsonic speeds would in any case heat). Don't need de laval nozzle theory here. NB look at most rocket plumes, they bulge out ie exit pressure is still above ambient. Small drag losses in a nozzle I suspect are just accounted for by an 'engineering factor' in practice.