The Am. Rifleman has set sort of a standard of using the average of five five shot groups in all their accuracy testing of new guns for the last several decades. Of course, that often isn't enough groups to find the difference between the two conditions being tested unless that difference found is fairly large.
There is no law that says you have to use the NRA standard. Most gunzines don't. Three shots groups make the product look better and pleases their advertisers. Even better they sometimes quote the best three shot group. That of course means almost nothing. Most of us test guns and loads to find answers not to make our guns or ourselves look better
Since only the two widest shots determine group size. The other shots are sort of wasted. (33% in a three shot group up to 92% in a 25 shot group) It seems to me the most efficient number of shots per group to get the most information out of our shots is clearly a series of two shot groups fired at multiple aiming points. Every shot counts.
The average of the two shot groups could then be converted to the equivalent size for 3, 5, 10, or 20 shot groups by known factors. Or as RicinYakima has suggested, finding actual composite groups by leaving up a backer target for the number of shots to find a ten or twenty shot group.
What do you think? For those folks that seem to shoot only a few shots in their testing the efficient two shot group would get them the most information for their money.
John