Not knowing why there was so much variation in .308 Winchester Factory Ammunition, I obtained 5 boxes from different lot numbers. The bullets were pulled and the powder weighed. One box had a 2.1 grain variation in the powder charges, and over the five lots there was a 3.5 grain variation with the 2.1 lot fitting in the middle.
That explained the accuracy problems so what do to next.
The powder appeared to be somthing like 4895 but I had a consistent source simply by pulling more bullets and salvaging the powder. An average of the powder weights, and a normally accepted accuracy load with Hornady match 168 grain bullets, produced consistent MOA accuracy or better in four rifles. One of the rifles had a 0.85 inch group at 200 yards on the first attempt.
While it may not have been a new discovery it did solve a problem of inconsistent groups.