Sorrow
Good post...and you mentioned a couple of things that can be added to.
First, let's look at a typical trim control layout...in this case, a P-51:
You make a good point about the size and location of the elevator trim wheel. As the picture shows, it is fairly large. It is mounted vertically to allow easier actuation. The pilot uses his left hand to rotate the trim wheel forward (clockwise) to trim nose down, and back for nose up. He can do this several ways.
Small adjustments are easy...just turn the wheel. For larger adjustments, many pilots found that they could make the input quicker by rotating the wheel with their palm...by placing their palm on the top of the wheel and then pushing or pulling to move the wheel through a larger rotation.
As you said, most trim inputs in a typical flight were made to the elevator. Rudder and aileron trim changes were less frequent. Because of this, these trim controls were twist knobs and were smaller and less easy to move in large increments.
While I seldom comment upon technique, I would note that arbitrarily adding 'nose up' trim on the back side of a loop would probably be unneccessary as long as the pilot intended on flying more or less of a symmetrical maneuver. If so, the exit speed would be about what the entry speed had been. Since the aircraft had been trimmed for the entry speed to begin with, the pilot would not notice any undue stick forces as he finished the maneuver. In fact, if he were to notice any out-of-trim forces during the maneuver, it would be at the top of the loop when the aircraft is at its slowest speed. If the pilot did not trim (nose up) in the pull up, the stick would get increasingly 'heavy' as his speed bled off...but then the feel would return towards 'neutral' as his speed increased coming back down.
With regard to A2A maneuvering, most trim changes comes as a result of airspeed changes brought about by maneuvers in the vertical or airspeed loss due to G-induced drag.
There are two factors in trimming during A2A maneuvering that should be considered. One is the magnitude of the trim change needed to keep the controls 'neutral' or 'hands off' as many pilots refer to it. In A2A combat, speed changes can be quite large, and the resultant trim change to maintain 'hands off' is also quite large.
Second is the time it takes the pilot to manipulate the trim control to make the needed change versus the time that elapses while he is in the maneuver. While the elevator trim wheel was designed to assist the pilot in making inputs quickly, often the speed changes occurred faster than the pilot could keep up with them. Often, the maneuver would be complete before the pilot would have the time to trim out the stick.
Because of this, many pilots elected to leave the trim alone during an engagement where speeds were changing rapidly. In prolonged conditions, such as a diving extension where the speed was expected to increase and then stay there, the pilot would probably trim off the stick pressure.
Finally, from a practical perspective, I found in my experience in A2A maneuvering that trim was way down my list of priorities. Not that trim wasn't important...it's just that other things had more of my attention. Things such as lift vector control, rates of closure, weapon envelopes, switchology, radio calls, SA on matter such as fuel state, fight location, and position of other bandits. These things kept me quite busy...trim was something I might adjust if I had nothing else more pressing...and this was seldom the case!
Andy