A bullet is seated long into an empty case. With bolt removed from the gun, the round is pushed into the chamber with a pinky finger pressing the case into the chamber as far as possible. Seating depth is reduced .002-.003 until the round falls freely from the chamber when the gun is tilted muzzle upward. So, essentially, you're finding that spot where the bullet no longer gets held in the rifling. Until that is found, each stuck round is lightly tapped out of the chamber with a cleaning rod.
In order to make this method as accurate as possible I would think one would need their brass sized and head spaced uniformly. However, there seems to be a universal standard of giving some length of bullet jump usually starting with .010 to .020. I can see this pinky procedure having an error of .002 with the jacket crushing ever so slight if one bullet were reused for all the measurements. However, I think it might be so slight to make no appreciable difference since it is common to adjust bullet seating depth in increments of at .010" to find that bughole grouping round.
I'd like to see some input on this. Pros and cons. Thanks!