No. 4 Rifle; Zeroing Instructions Data Inconsistency?

[Surpmil]

No 4 RIFLE TESTING
For the No 4 Rifle, the accuracy test was the same at 100ft ten per cent of all rifles were then fired at 200 yds when six of seven shots had to fall in a rectangle(sic) 6in x 6in , the point of mean impact having to be within 3 inches of the point of aim in any direction. Ten per cent of rifles fired at 200 yds were again fired at 600 yds when 6 out of seven shots had to be in a rectangle 18 inches x 18 inches the permissible deviation of point of mean impact being 9 inches up or down, or left or right. Two per cent of rifles were fired from the shoulder, ten rounds being fed into the magazine by charger and fired rapid to test “feeding up” and ejection. After these tests the barrel was inspected to ensure that there was no expansion in the bore or chamber and that it shaded correctly from end to end. (Was not bent)

When describing "six of seven shots had to fall in a rectangle(sic) 6in x6in, the point of mean [main] impact having to be within 3 inches of the point of aim in any direction" the question would be how was "mean/main point of impact defined?

The edges of a 6x6 square are always 3 inches in any direction from the center*, so any such mention of a "mean/main point of impact" is completely superfluous unless out of that six or seven shots a certain number were required to be more closely grouped than 6 inches(?)

Or were they just saying the same thing twice to better ensure it was (mis)understood??

Regardless, the 18x18 inch acceptable dispersion at 600 yards, effectively 3 minutes of angle, implies pretty clearly that many rifles, and whatever ammo was used to at least test them, were not of a high enough standard of accuracy to make the 2 inches at 100 yards under discussion here anything that could be relied on "in real life".

A guess could also be hazarded that if this 8.5" related to anything else, it might be that it was thought preferable in zeroing practice and perhaps in combat to have the actual fall of shot slightly higher than the POA?

* We'll leave out the matter of the distance to the corners, and why they stipulated a square and not a circle, as though a greater error in those directions than at the vertical and horizontal axes was somehow of less import !