Well, if all of your other variables are all over the place, why would you care about one that, more than likely, is pretty far down on the list of significance?
Not to mention, unless you are looking to change everything you do, you would only alter the one and leave the rest of your routine “as is”. Now all you have to do is see if the one variable, seating depth variation, makes a difference on paper. Doesn’t matter much if other variables, cancel out or negate any added gain or not, if you are just trying to understand if elimination of slight seating depth variation makes your process better, worse or makes no difference at all, on target.
Then you can either check that box as “important” or scratch it off your list as a path that needs to be followed.
I suppose you could jump through all the hoops and change everything at once but then you won’t know what variables are key and what ones you didn’t have to mess with in the first place.