Straight comparisons based on group sizes are misleading as changing the primer affects the powder burn characteristics, hence group size. The load may have to be fine-tuned in effect to suit the primer, and even then the results would only apply to that single combination of cartridge, bullet weight/model, and powder make/grade.
There are so many variables involved in this that no sane person would wish to take it on board even for a single cartridge. What can be shown is the effect on internal ballistics for a single cartridge / load combination / rifle and different primers affect both the average MV value and the ES / SD values. They may have little real relevance to the short-range competitor as some combinations with terrible ES values can also provide the smallest average group size in a particular rifle. They are of great interest to the L-R competitor where large velocity spreads = vertical dispersion on the target. This is particularly so where a single 'recipe' is used, a good example being Palma shooting (.308 Win of course) where the US teams used to load Winchester brass, F210M, 155gn Sierra MK, and 40-something grains H. VarGet. Lapua was requested to produce small primer / small dia. flash-hole match brass and this reduced ES values by ~ a third, a great success now used by the teams exclusively and Lapua putting its 'Palma' brand brass into production.
I've recently tested 16 LR and LRM primers in 308 Win in a single FTR rifle using 100 RWS cases, 168gn Hornady HPBT Match bullets, 44.0gn Viht N140 from a single production lot. 17 round batches were loaded to plus or minus 0.04gn charge variance using Acculab scales, two rounds fired as foulers starting with a clean barrel, then 15 MVs recorded.
Average MVs covered a 35 fps range from 'mildest' to 'hottest' primer (2,780 fps lowest / 2,815 fps highest) and ES/SD values ranged from excellent to dire with the best barely in double figure ES and the worst >40 fps. The average of three 5-round 100yd groups varied, but not by a great amount, and bore NO relationship to the primer's ballistic consistency, the worst / hottest performer producing the smallest average group value!
This was not a 'scientific' test in any way. To do that would have required far larger samples, an indoor temperature controlled environment, selected and fully prepped brass re-annealed after every firing, (or only new brass from a single production lot used for a single test firing) and likely two new identical barrel blanks with identical chambering swapped after a certain round count!
Even then, it would only count for 308 Win with the 168gn Hornady and Viht N140 in a particular spec match barrel with a single chamber and throat configuration. Changing the powder to say VarGet and working a new load up for it might or might not change relative primer rankings. (Actually, likely not too much - but swapping .300 WSM or RUM for 308 Win would be another matter entirely!)
It did show that primer choice does matter (quelle surprise!), that generally magnum primers aren't a great idea in a .308 match load (but not that they always give high MVs), that CCI-BR2 and Fed 210M primers (in my lots at any rate) performed significantly better than their F210 and CCI-200 standard equivalents.
Come next summer, I'll do it all again using Lapua Palma brass and Small Rifle primers, but likely with a different bullet -powder combination, as this is being done on the cheap with what would otherwise be spare barrel life and cartridge components. It will definitely be a different rifle as the 308 Win job that was used will be 7mm-08 in a matter of weeks.
