Yet another, "back-in-the-day" trekk: when Dave Tooley and [the late] Charles Ellertson talked me into making thirty Cal. FB bullets (187 Gr.) for use during IBS 1K competition, the bullets proved
exceptionally good - they were well suited to 1:12" to 1:13:" twist rates (@ Std conditions, the latter delivering Sg 1.4+).
Some of the users were using the Jeunke (Sp?) device to attempt sorting various brands & bullet types - mostly, "VLD" (secant ogive/long BT) type bullets. Most, who sorted via this device, considered anything over 5 deviations useful for fire-forming/practice only . . . I cannot recall all of the deviation details. A good percentage of the BT bullets were deemed unworthy of tournament use.
The 187 Gr. BIB bullets (J4 1.300" long - always 0.0003" or less wall variation) always sorted into
< 4 deviations - so, naturally, the fours were sighters and fire-formers . . . yep humans are a weird breed.

People using those FB bullets did remarkably well - especially when one compared OPPORTUNITY . . . Perhaps Dave Tooley will chime in and provide some of the stuff I have forgotten.
During that time, on several occasions, to different individuals, I sent sample packs, with jacket wall thickness varying from exceptionally good (
<0.0001" wall thickness variation) to horrifically poor - up to 0.0015" variation.
Only yours truly knew which of the coded bullets were made with which ("quality") jackets:
not a single Jeunke user came anywhere near to [correctly] sorting based upon variation/amount of jacket wall-thickness. I do not know what the device measures, but firmly believe it is NOT wall-thickness uniformity. What is THE standard for calibration - what IS a deviation? Everyone I was involved with used a few bullets - what if THOSE were, "bad"?? Dark science . . .
Several individuals were so irritated by my test and belief, base upon their Jeunke results, that they quit talking to me and buying bullets!

Yes, the packaging was, "tricky" each baggie contained a single bullet, and there were varying combinations of wall variation, with duplicates . . .
again, not a single individual came close - nor, did their results correlate.
For cup & core bullets (relative specific gravity, lead:copper), we can determine that center-of-gravity off-set is about 30% of the wall-thickness variation (rounded up to 1/3rd). Thus, if the bullet maker is properly QCing the jackets (knows the correct DATUM points and tolerances), and is either using J4, or, Sierra Match King (specified maximum wall variation 0.0003" at first DATUM above the inside base), the MAXIMUM center-of-gravity off-set would be about 0.0001" . . . certainly not much.
I sure hope Dave will come to my rescue!

RG