Grains avoirdupois is a measure of weight.
Where I think the divergence here is arriving is more a question of powder measure consistency and further a question of how precisely one must measure black powder to see an effect on the target. I posit that where in smaller charges a tenth of a grain variance of a given smokeless powder may produce marked effects on the target, a tenth of a grain charge weight
alone of 2f or even 3f will not.
I arrive at this conclusion from the point of departure that in a 357 magnum cartridge (used as an example) one uses 25 grains of Goex 3f under a 158 grain bullet to get a velocity of 900 feet per second with a 158 grain bullet, according to Goex website load data. 900/25=36. With a 158 grain bullet using 4.5 grains of Unique one gets 859 fps. from the Alliant website. 859/4.5=190.8. This order of magnitude of sensitivity as a function of velocity of propellant is significant.
Ferinstance, 25.2 grains of 3f multiplied by this constant still gives 907 fps, for a difference of 7 fps, where 4.7 of Unique gives 896 fps. for a difference of 37 fps. --I recognize the shortfalls of this analysis, but it illustrates a point. I would think that the natural spread of velocities inherent in the mix even with weighed charges of propellant would exceed 7 fps, and I doubt that 7 fps will make a difference in shot placement on a target at pistol distances.
If one multiplies a couple of charge weights by these constants one can get an idea of how much a couple tenths of a grain difference in charge weight will affect the resultant velocities, and small velocity spread is important to accuracy. The smokeless powder is much more sentitive to a variance in charge weight than black powder, as a function of less mass of propellant provides equivalent energy and velocity.
I posit that at some point of variance in charge weight of black powder one will see a difference in performance, but that black powder can accept a wider range of charge weight variance than can smokeless and still perform the same, as far as human abilities to perceive it can measure or evaluate.
Of greater importance perhaps than a charge weight variance of less than .1 grains with black powder is its consistency. I've seen fellas sifting their powder to assure uniformity in the powder granules they load, and claim that it does make a difference. The manner of pouring the powder will change the load density, and variance in compression will affect its performance more than the 10th of a grain weight variance, or even several tenths of a grain variance.
So whether one weighs each charge of black to the tenth of a grain, or whether one has a powder thrower set to volumetrically throw a charge of the same weight within a tenth of a grain, it would seem to matter less than how the charge is introduced to the chamber, and how the ball is rammed home. Where with smokeless, a closer tolerance of charge weights will generate a greater uniformity of performance, because one does not typically compress smokeless, which tends to perform better with a little air in the case.
I cannot dispute that the high scorers in black powder competition weigh their charges, but I doubt that such a method gives higher scores. It is difficult for me to accept given the propellants nature that this technique alone results in higher scores. I suspect that sifting the powder to remove fines and charging their cylinders in a uniform and consistent manner, and ramming uniform balls in a uniform manner does more for them, as well as installing caps in a uniform manner, fouling mitigation technique... plus other factors I haven't considered. --Plus that those guys are damn fine shots by any standard can't hurt.
Perhaps the weighing of charges is a red herring introduced by those hard holders to distract the unwary from the techniques that are really producing the results?