Okay, I should have specified what I am thinking about. My bad.
Let's say, for example, I have agitation method 1 and method 2. They are wildly different, in some way (timing, soft versus hard inversions, what have you).
What I am envisioning is taking two rolls out, setting the camera up on a tripod, and shooting two or three different scenes, by rotating the camera, but keeping the tripod in the same place. I take five to thirteen shots of the scene, Over and under-exposing by two stops, either in full, half, or third stop increments. Then I switch to the other scene, and repeat until I fill up the roll, maybe with a lens cover frame between the sets, maybe not.
Then I load the second roll and repeat the exposures on the same framing, same shots.
When I get back to the darkroom, I develop each roll separately, changing nothing but the agitation method between the two.
Then I pick one of the frames from roll one and get it to a split filter test print. Using the same frame from the second roll, I use the times for the first test print, then compare the results of the two prints, to see if there is any difference in the highlights or midtones. I don't suspect there would be much of an appreciable difference in the shadows, unless Method 1 was, for some reason, not to agitate at all.
Has anyone ever done this, either for agitation, or change in developer, or some other single change in process?
Bayard
This is an interesting thought process, but I don't know whether I would undertake the work!
I can see doing something like this with an automated processor that allows you to adjust variables like rotation speed, but otherwise I'm not sure that it is possible to assure sufficient repeat-ability to reach the conclusions you want to be able to reach.
In addition, I think you would have to do a pile of testing in order to adjust your times so as to match contrast between gently agitated development and vigorously agitated development. Otherwise, the print tests would be misleading.
As I said, the thought process is interesting, and I think valuable as well. It drives home where the variables might be found, and makes it easier to systemize your approach in order to ensure consistency.
FWIW, I have personally standardized on:
1) Initial, single, vigorous inversion with twist, followed by a bubble dislodging "thump";
2) place tank on rotary, auto rotating and reversing agitator, start rotation and timer;
3) after 30 seconds remove tank from rotary agitator and stand tank upright;
4) complete development by starting at the one minute mark and repeating every 30 seconds thereafter by giving the tank 5 seconds of twisting and inverting randomized agitation plus "thump".
Each 5 second agitation includes two full inversion and twisting cycles, with each cycle being sufficiently vigorous as to cause the developer to "tumble" through the film.
For each of the following stop bath, fixer, rinse and HCA steps that follow I do one hand agitation cycle at the beginning and let the the auto rotating agitator do the rest of the work. The fixer stage is broken into two equal steps with two batches of fixer - I set the time by doing a clearing snip test each time.
Mostly, I'm using Paterson tanks, with plastic AP/Arista/Samigon reels. I don't use the rotary agitator for the full development, because the film moves around in the reels too much in the development stage. That problem doesn't seem to occur during the later stages.
....I 've just remembered my very 
.
.
