That's interesting, can you explain why?
I'm not attempting to directly answer your question, just offering one perspective on powder weight variance, or the amount of error we might routinely experience when weighing powder. As an example, I have three F-TR rifles that are set up similarly, and tuned loads with each almost routinely consist of 43.5 gr Varget under a 200.20X bullet at ~.009" to .012" off the lands in Lapua Palma brass with a Fed 205 primer. Such a load typically generates ~2650 fps from a 30" barrel.
In this scenario, 2650 fps/43.5 gr of Varget means roughly 61 fps velocity per grain of Varget, or just over 6 fps velocity per 0.1 gr Varget. I understand the relationship between charge weight and velocity is not always perfectly linear, but we're making that assumption for use in this example. What this kind of simple calculation does is allows one to get a rough idea how much a given charge weight variance or error might be worth in terms of muzzle velocity. This is useful because then one can begin to formulate some idea of how much charge weight variance they must be willing to accept for a given velocity variance. In this example, weighing powder to +/- 0.1 gr Varget should be worth roughly +/- 6 fps velocity variance solely due to charge weight variance.
Such calculations are far from "perfect"due to the assumption of linearity made above, as well as other caveats associated with the reloading process. Nonetheless, they can give us estimates that are simple to make, and that we can use to try and understand what our limiting sources of error in the reloading process might be. What I mean is that if one weighs powder to within +/- 0.1 gr Varget in the above example, which in theory should roughly correlate to +/- 6 fps velocity variance, but one actually then observes a velocity extreme spread of 30 fps or more during testing, it is likely
not all due to charge weight variance. In other words, there are likely
additional sources of error involved. This is critical because if the observed/measured velocity variance is far larger than the theoretical value estimated from charge weight variance, then weighing power to a much finer increment is unlikely to solve the problem.
Likewise, if one has some idea what their minimal acceptable velocity ES value might be, they can easily make an estimate of how fine a powder weight increment (i.e. precision) they need to use, which might directly affect the method they choose to weigh powder; i.e. how precise/accurate does their scale of choice need to be. Per the above example, if I wished to have a theoretical velocity variance of no more than +/- 5 fps, I would need to have a charge weight variance of no greater than approximately +/- .08 gr of Varget. Bear in mind that one kernel of Varget (on average) weighs about 0.02 gr. Thus, there are
approximately 5 kernels of Varget per 0.1 gr charge increment, and from the above velocity per charge weight estimate, one kernel of Varget might be worth about 1 fps velocity, at least in theory. This is just to give you some idea of the relative scale we're dealing with here. Again, these numbers are not written in stone, they are simply estimates one can readily make to gain some insight on potential sources of error, such as to what level of precision their powder weighing setup needs to work to keep velocity variance due to charge weight variance below some level.
There is one final consideration in terms of sources of error. What we really wish to identify first are the
limiting sources of error, or largest sources of error, because minimizing those will have the greatest effect in term of generating consistent velocity in our handloads. In fact, if we can minimize a particular source of error to a sufficiently low level (i.e. far below any other sources of error), it effectively ceases to be a variable. In the context of this discussion, if one can precisely weigh powder to an increment predicted to cause well under 1 fps velocity variance, a level of accuracy that most chronographs cannot achieve, we can
effectively eliminate charge weight variance as a source of velocity variance. That is important for at least two reasons. First, those that weigh powder to +/- half a kernel (or less) really don't ever have to worry about charge weight variance as a contributing factor to unacceptable velocity variance when they're behind the rifle in a match, or just working up a load. In other words, you can effectively remove charge weight variance from the reloading equation. Second, if they encounter unacceptable velocity variance during the load workup, they can be pretty certain it was not caused by excessive charge weight variance. Thus, they don't have to waste time troubleshooting their charge weighing apparatus, and can focus elsewhere to identify the cause. Attempting to effectively eliminate charge weight variance as a limiting source of error is likely one of the major reasons you might find some competitors weighing powder to a much finer increment than that for which they could ever
realistically shoot the difference. It's not done with the notion that somehow their scores will go up if they weigh powder to an even finer increment. Rather, by weighing powder to an extremely fine increment, they don't ever have to worry about it becoming a limiting source of error.
Again, I am not making any suggestions here as to what level of precision anyone else should be using for charge weight in their chosen shooting disciplines, or what methods they should be using to weigh powder. I am merely offering a simple method that they might use to help make that decision on their own. Understanding and identifying limiting sources of error is an important part of the reloading and shooting process. One needs to have some grasp of this concept so that they can choose a suitable method for weighing powder that fits their needs and their budget.