I'm pretty late to this discussion, but I've found the following article interesting. It discusses differing number of shots/group and the relationship between them with respect to group size. The graph near the end of the article and the table giving correction factors for differing shots/group are particularly useful. This latter table allows you to estimate the group size for a larger number of shots than you fired--e.g., estimated 10-shot group based on your average of several 3-shot groups.
http://www.the-long-family.com/Group size statistical analysis.pdf
As for my own bench shooting and load-development, I use the NRA standard 5 x 5 system--an average of 5 consecutive 5-shot groups at 100 yards. The "consecutive" part of this is important, as it rules out cherry-picking the best groups in arriving at the average. Instead of estimating the MOA potential of my rifles, I find this average of 5, 5-shot groups the most meaningful index. Too many shooters declare their rifle as a "1/2-inch" gun on the basis of one exceptional 1/2-inch group.
http://www.the-long-family.com/Group size statistical analysis.pdf
As for my own bench shooting and load-development, I use the NRA standard 5 x 5 system--an average of 5 consecutive 5-shot groups at 100 yards. The "consecutive" part of this is important, as it rules out cherry-picking the best groups in arriving at the average. Instead of estimating the MOA potential of my rifles, I find this average of 5, 5-shot groups the most meaningful index. Too many shooters declare their rifle as a "1/2-inch" gun on the basis of one exceptional 1/2-inch group.