i can't tell you how much it affects performance. I guess that would depend on what sort of performance you're looking for.
I can give you an example of one of my precision loads. This is for the 6xc caliber using Norma brass.
I bought 300 rnds of it and found the average weight of a large sample (probably around 25 cases chosen at random) was 158.17g with a stddev of .47g
I found many of the cases fell into groups, so I made 3 "lots" of brass.
lot 1:
avg weight 157.99g
stddevp .07g
max weight 158.08g
min weight 157.88g
lot 2
max weight 158.48
min weight 158.26
lot 3
max weight 158.69
min weight 158.49
so basically, cases in each lot would vary by no more than two tenths of a grain.
This level of selectivity resulted in roughly half of my brass being discarded. i.e. out of 300 cases, about 150 fell into one of those 3 lots.
generally, I've found my lapua brass to be a little more consistent (unfortunately, they don't make brass in 6xc) and winchester brass to be considerably less consistent.