Here's what I do ... to find the 'Node'.
GIVEN -
Every barrel on any type of firearm or firelock when shot, whips in a sine wave manner, vibrating like a sideways 'S'. The node is where the barrel is at the point(s) of the LEAST movement. That is when you want your shot to be fired. For some of us, good enough is
good enough, but I do find this very interesting and fun to experiment with! And yes, it works with muzzleloaders like my 1560 wheellock just as well as it works on my precision 1,000 yard toy that goes bang.
I would repeat your testing and would test 3 to 5-shots each benched, starting @ 30, 1/2-way to 35 or 32-1/2, 35, 37-1/2 and 40-grains.
All on the same target, like shown below. This allows you to compare group sizes. Sometimes 2 powdah charged
adjacent to each other will display the smallest group (ignore where they impact - shoot for group size!). That would then be the charge range people further experiment around.
One fine point, if you truly want to eliminate YOU as the source of the variation, but clearly this takes some thinking, planning and careful loading. Do not fire the 30-grain charge all as one group, but VARY what load you shoot when, always switching it up. I make a simple table of the charge and number of the shot (like 1 thru 3) and then work around the table until I have fired 3 (or 5) shots of each at each respective bullseye. Statistically what this does is to normalize the data so that you can rely on it, significantly increasing the confidence you can place in the results.
Now please be aware that many will look at all this and say,
"Harummphhhh, none of that is needed to have good results with black powdah ... " and they can be right. But -
you asked - and this is the best way to wring all the precision out of your fire lock that you can. But look again at my target above ... which load would YOU use?