I wonder if that would be true at steady-state (e.g. when the train is effectively empty), but for a train at rush hour, whether much of the capacity of the AC

I wonder if that would be true at steady-state (e.g. when the train is effectively empty), but for a train at rush hour, whether much of the capacity of the AC system is already being used to remove heat generated from the humans, and if that means the temperature difference to the outside is actually mostly unimportant. That is to say, the person next to you is generating and dissipating a fixed rate of heating onto you (equal to their basal metabolic rate consumption), so all that the jacket can have changed is their steady-state perception of the local temperature.