Waiver...I am NO expert on stringology ;^)))
I will share a few pics/thoughts just the same....
1st , I fully agree with what thoughts Kirk has shared as well. Much logic ...
It seems to me over the great number of years I've been purchasing and then making strings (flemish twist) , that a lot of the "issues" arise from uneven pressure/force while twisting them up .
1)...The picture added, shows "on the left" a string purchased from one of the Nations Best . The one on the right (Red&Black) is self rolled/twisted;) . You can see the one on the left has one side slightly larger due to not twisting as evenly with the same amount of force/pressure. Though crafted beautifully & it has performed beautifully...None the less you can see a difference.
2)...I also think a lot of these issues can be corrected by simply "overloading" , pre-stretching the strings to "settle them in" . I did learn to do that from another chap on Youtuby;)
Having a background in load testing , it was standard procedure to overload the load cells as part of the testing. It helps to "stabilize" the material. The load cells I tested/calibrated were of various metals....and I know some will say "apples to Oranges' ' ;^))) The same basic principles apply within a testing perimeter .
I usually use 200#-300# of weight/force to preload/overload my strings (pre serving)which settles them in rather nicely with No change in brace/length after doing so. Stabilization
3)..and lastly , my thoughts /findings are ...keep your pressure , amount of twist duplicated as close as possible, and preload/overload those strings to settle them in physically and structurally ....As the fibers in the material at hand (b-50-b-55-fastflight-D-97-Rhino)...all of them will differ to a degree roll to roll , inch by inch , fiber by fiber..