I second. That has been my experience exactly. Believe it or else, paper is tougher than lead under compression. The paper gets squished just so far and then it sizes the boolit. I believe a steel sizing die will be more consistent than a paper one dont you? Unwrap about 5-10 boolits and measure the diameter that the lead was sized down to. If they are being sized below bore diameter, switch to a thinner paper. If they are still over bore diameter, pick the smallest one and size them all to that size before patching them. Roll them under a file to "knurl" them, then patch and size.If you unwrap a sized pp bullet and measure with a mike it'll tell you what your core diameter is now.
By doing that and testing I found that even with my rough home hand made sizing die it halved at least the group size.
The cores ended up different diameters all up and down the core from wrapping then sizing only.
Can't be good when you start to spin them fast and not expect them not to go whizzing around in a big circle instead. Althou how consistent it might be.
Another thing you can do is unroll 5-10 of your patched boolits and see how consistent the finished patch is in height. When I did this, I got all different heights of paper and extremely jagged bottoms from clipping off the tails. I started leaving a dot of exposed lead on the base of the boolits by rolling the edge of the paper over the base of the boolit, like 303guy does. then using my special tool to trim them to length. this created a much more consistent patch.