Wiljen
I see where I confused the issue; They use the reference ammuntion to get an "offset" figure to then add to their test loads. That sentence is misleading; should have said they use the reference ammuntion to establish, with the offset, what is the expected psi of the reference ammuntion in their test barrel. That then establishes the base line for load development and a % is, or at least should be, used with additional load development. We are quite in agreement there as I also use a % rather than the set figure to adjust the psi at all levels with load development. However, with the 7.5 gr Unique load of this thread the actual psi's are as recorded. The test barrel I used gives psi's, as noted, within the expectied variation from the reference factory ammuntion using that same lot of ammuntion.
I'm just reiterating what is industry standard according to SAAMI instructions. The above is the minimum they apparently are comfortable with. I, as with you, am not comfortable with just the one offset. I use a minimum of 3 myself using different factory ammunitions as reference ammuntion. SAAMI also is comfortable with one 10 shot string. I am not. I prefer a minimum of 3 ten shot strings to verify accuracy and psi with the highest level of confidence. Most of the ballisticians are comfortable with a minimum of 7 shots in a string for "confidence" that a load is correct. Many use the SAAMI standard of ten shots though. You might have noted in my discussions with DrB that he disagrees and feels a 5 shot group is good enough so even educated opinions vary.
I still prefer 3 five shot test strings or 3 ten shot test strings for psi "confidence" and 3 ten shot test strings at 100 yards or one ten shot string at 100 yards with another ten shot test string at 200 yards to have "confidence" in the accuracy of any load. I have come to prefer those sizes of test strings over the years becuase I have been burned by too many 3 and 5 shot groups, velocity averages and psi measurements that proved to be not correct. Most ballisticians and reloaders find that many in a test to stretch their resources of supplies, $s and time so the lessor numbers are most often used. As mentioned though, for absolute confidence I use 3 ten shot strings for accuracy, velocity and psi measurements but will use 7 - 10 shot test strings for intial work up or for data gathering.
My appologies for the confusion, I have added an additional explanation to my post above.
Larry Gibson