First, some of what we are taught from "way back when" is bastardized in our modern world. When you get something like, "tilt the lock a bit down and away from the barrel to get faster ingnition", and we find that wrong, the question should be "why did they think that?". The experiment is with very pricise testing equipment, NOT with an actual lock on an actual gun barrel, SO perhaps that bit of advise was from a time when the touch hole would easily have been plugged by the priming powder jostling around inside the pan, under the closed frizzen, as the shooter or hunter walked around, and it would pile up against and covering the touch hole???
As for it mattering..., well his test only shows what happens between the pan ignition and the ignition of the powder in the chamber. The bullet doesn't exit in .032 seconds. The charge merely begins its work in .032 seconds. The next question might be..., how much time delay before the bullet begins to move from its resting position...., add that to the .032 seconds, and THEN..... how long before the bullet exits the muzzle..., giving you total time from flash to exit, which then would tell you how much lag time for human error to cause a change in point of aim.
Is it possible that part of the reason the patched round ball is often more accurate than its heavier, conical cousin, because the lower mass of the patched round ball makes the ball to begin moving sooner, so lowers the time required to exit the barrel, and lessens the time for human error to effect the aim of the barrel??
LD