@Dave M.
Sort until you have an obvious normal distribution of weights.
Take 10 low outliers and 10 high outliers, and the rest from NEAR the mean.
If outliers weighed to the nearest milligram don't show any benefit, weighing to 0.1mg surely won't.
Load your 70 or so rounds for your 600X3 isolating the ten low and 10 high for your second relay.
Ex: Sighters from mean weight, relay one from mean weight, first 10 of relay 2 low outliers, second 10 of relay 2 the high outliers.  Shoot relay 3 from the mean weight primers.
All same day, almost the same environment, same reloading batch, same you 

Post scores.
One hint for those without a precision lab balance, or just to save a few seconds per primer:
Tare the scale then add a 1 gram check weight to the scale.
Will indicate 1.000g without a primer and 1.350g (or so) with a primer.
Milligram resolution should show outliers.  With a 1g base, DRIFT and REPEATABILITY will be instantly visible.
Auto zero of balance won't be triggered.  At a very small measuring range (light to heavy primer) scale LINEARITY will NOT be a factor.  Full scale calibration, percentage based, will NOT be a factor except in the WORST conditions.  The 1.000g will monitor every weight performed.  If it drifts, remove the check weight, tare, add the check weight.
This 1.000g check weight is a continuous IN PROCESS check.
If you don't trust the process, reweigh the lowest and heaviest outlier after sorting.
Red and green markers in the case groove to track outliers in case you spill the box