I plan to shoot 300 yards and more. I have brass that I've resized but never fireformed. I plan to carefully trim them to the same length and find the most consistent volume among them. Again, I hope volume sorting works. My concern then is whether the energy used to expand these cases will affect accuracy compared to my current batch that are fireformed and neck sized.
I always throw a charge under my desired load and trickle up to my desired powder
weight.
Several items:
1) The primer pocket can be plugged simply by inserting a spent primer upside-down with a priming tool. Alternatively, you can buy [relatively] inexpensive and re-usable rubber plugs designed for this purpose. Either of these approaches allows the case to stand upright (i.e. flat bottom), which makes water volume determination much simpler.
2) The energy necessary to expand virgin brass is not inconsequential. Reproducing the same velocity from a load worked up in virgin brass using fire-formed brass typically takes anywhere from 0.1 to about 0.5 gr
less powder. Fully prepped virgin and fire-formed brass are totally different animals. Does this affect "accuracy" (i.e. precision)? I'll answer this with another question, "Does velocity affect precision?", because virgin versus fire-formed brass can most definitely affect velocity.
3) From your posts, I am receiving the impression that you're using mixed brass (i.e. you mentioned sorting brass by headstamp). Is this correct? Perhaps a better question is what is the source of the brass you're prepping? If you have purchased different Lot#s of brass, even from the same manufacturer, keep the different Lot#s separate, and only work within a single Lot# when doing your sorting approaches.
If this is range brass you picked up (as I mentioned, I can't really tell from your posts), you
may be able to improve the consistency of internal volume via sorting by headstamp, then determining actual water volume. However, there are no guarantees that the the case wall thickness in every region will be exactly the same among pieces of mixed brass, even if you sort them by volume first. It also means that sorting mixed cases by weight as a surrogate to volume will not be a fruitful endeavor. Other than sorting by headstamp, I would not spend the effort trying to sort mixed range brass by volume. There are too many other variables involved for that to be a useful approach.
4) Chances are pretty good that if you think up some test you wish to try, someone here has already done it. So you're doing the right thing by asking these questions. Your question regarding the weight variance between charges thrown with a machine alone, versus trickled up by hand is well-documented. Powder throwers typically have an innate level of precision (i.e. thrown charges will typically weigh XXX +/- some amount). Part of their precision depends on the powder you're using, as the size/shape of the kernels can affect the thrower's precision. Nonetheless, the precision achieved with a powder thrower alone will rarely be nearly as good as what you can achieve by hand, setting the thrown charge weight slightly low and then trickling up by hand, assuming the scale/balance used has sufficient resolution.
The bottom line is that once we have found an optimal load, we always strive to make every single loaded round
exactly the same. However, there are always sources of error (sources of variance) in every aspect of the reloading process. The key is to identify the
largest sources of error and deal with them first. Smaller, or non-limiting sources of error can sometimes be cumulative, but they are also generally more difficult to remedy than the largest sources of error.
My point here is that attempting to deal first with a smaller source of error when there is a [much] larger source of error present will generally show little effect. Variance in case volume can certainly be a concern for precision, but it generally takes more case volume variance than you might expect to have a noticeable effect on velocity/pressure. One way to test this is to make two sets of 5-10 cases each; one set that all reside at the lower end of the case volume range, and one set that all reside near the upper end of the range. Load the two sets with an
identical charge weight, then determine average velocity with each set. That is a relatively easy experiment to determine prior to expending any great sorting effort whether the range of case volumes you're using actually cause a noticeable effect on velocity with a given charge weight. Only if the answer to that question is "yes" will sorting cases by volume have a significant effect on velocity variance.
Perhaps the hardest part of all this for reloaders is to actually identify the major sources of error in their reloading process so that those can be tackled first. The second hardest part might be deciding exactly which steps you wish to incorporate in the reloading procedure, and which are too time-consuming, or provide minimal benefit for the time required. Unfortunately, dealing with second hardest part really requires the hardest part to be addressed first (i.e. what are the major sources of error?). I have a feeling from the direction of this thread that it might benefit you to describe in a little more detail the setups you're using (both rifle and reloading setups), the reloading components you're using (i.e. source of brass, bullet type, powder, etc.). That way, responders might be able to give you a better feel for the steps you should be taking, and the steps that are too labor-intensive and are unlikely to buy much better precision. Just a suggestion.