One can certainly "ballpark" an estimate, and it will likely work pretty well. The downside to that is that making a good estimate usually requires empirically determining velocity with a given charge weight in a few different case volumes so that the estimate of how velocity changes with respect to case volume can actually be made. Alternatively, reloading programs such as QuickLoad and GRT can make these predictions given solid inputs. However, one problem with any of these approaches is that they're typically based on the assumption that the effect of chagning case volume must always be a linear response. I can assure you that that is not always the case.
An example of this is that combustion of powder is a chemical reaction. We use a defined amount of reactant (i.e. powder), in a defined reaction chamber or pressure cell volume (i.e. the case expanded to match the chamber space), at a given temperature, so on and so forth. The problem lies in that the chemical reaction (i.e. powder burn rate) will increase as pressure increases. Thus, as pressure increases due to increased charge weight or decreased pressure cell volume, the powder burn rate can also increase, leading to pressure increasing at a rate even faster than would be predicted from a simple linear response to charge weight or a change in case volume. In other words, the response is a "curve", not a straight line. This effect will be even more noticeable as one approaches MAX pressure. As long as one uses previous experimental data or one of the reloading programs to make predictions within a certain range of charge weights or case volumes, the "curve" will effectively be close enought to linear within that limited range that the estimates are pretty good. The farther away one tries to predict such changes, usually the farther away from "reality" the prediction will be.
So a lot of this depends on how much effort you want to put into this. In your position, I might consider conducting the following test: sort your brass into three volume groups - low/medium/high. Keep track of the values you used to sort the cases into each group (i.e. the group sorting parameters). Load up a few rounds of each holding all else besides case volume as close as possible (i.e. use the same primer, charge weight, bullet, seating depth). Then shoot each group and record velocity. With only minimal effort, you should be able to generate your own "formula" based on the velocity versus case volume data that will provide a reasonable estimate within a case volume range that is wide enough to be useful for you as you move forward with the reloading process.
