Please don't confuse sensitivity with accuracy. You can get scales which are sensitive enough to consistently weigh a pellet (and you should try weighing the same pellet 10x just to see if there is variability (that makes a difference)) but if you want that number to be accurate as well then you will pay a lot more (consumer legislation, etc).
Best thing is to pick a tin from a batch and have 5 dishes. Just decide what range you think is worth using (e.g. JSB Exact say 0.547g so I would use 0.545 to 0.550 as the middle dish, either side 0.530 - 0.544 and 0.551 - 0.565). You will soon get a feel for the range in weights as most should be in the middle dish, a reasonable number either side and few (if any) in the real outliers - way too light or too heavy. The ranges will vary depending on the accuracy of the scales (you pay more!) but you can adjust your 'bin' ranges with less accurate scales - you just need consistency (and frequent re-Tare-ing as the zero drifts). Even your body heat will affect sensitive scales.
No need to count them, just weigh the groups and divide by your 'centre' weight.
Bin the real outliers but then shove the middle 3 dishes back in the tin as they are all near enough (weight differences being in the 'length' of a pellet as they are nibbles of wire and not molten lead!
You soon get a feel for a batch quite quickly and there is no need to weigh them all. If there is an awful asymmetry then maybe try another batch :-)
Oh and thee outliers work fine - the consistency is really only needed beyond 45yds.
Very tedious to these things properly but anyone who has worked in a lab knows such things are worth doing consistently.
BFTA Secretary 2012-2014
CSFTA Secretary 2014-2016