It's a waste of pellets for me to keep going through the same hole so I stop after 5
For me, if the first three are nicely shot, the last two are a real heart thumper to try and keep a nice grouping. adding more shots to it would likely cause me to twitch as i shot since i am more worried about keeping a tight group than on concentrating on the individual shot. Also, with fliers very likely, after 5 good shots I hate to ruin a good thing (on the RARE occasion I can get a good grouping) ;- )
From a statistical perspective, I usually choose a 95% confidence interval for significance. Anything less than an N of 5 (5 shots) and I feel that there isn't quite enough of a sample population to give a good prediction of variance. N = 10 is commonly accepted as a reasonable proxy for an infinite population size. I. E. t = 3.182 at N=3, 2.571 at N=5, 2.228 at N=10, and 1.96 at infinity at the 95% confidence interval.