After some back-and-forth with The Evil Boolit Brothers (Goodsteel and Sgt. Mike) we devised small test to try to quantify the effect that an imperfect cast bullet would have in a high velocity cartridge. We figured that Big Bertha, my XCB target rifle, would make a suitable test bench, and I prepared the test as follows:
20 bullets, NOE 165 gr. XCB variant, cast from Rotometals Hardball, visually inspected and weight sorted to within 0.1 grs, were selected. 10 of them were gas checked, lubed with Lars 2500+ and sized to .309, and then loaded over 45 grs. IMR4831. This load had given reasonable accuracy before, and had chronographed in the 2550 fps range, so no chronograph was used for this test.
The other 10 bullets were prepared like this:
A 1/16" hole was drilled slightly off center in each bullet base, about 1/4" deep, to simulate an internal void in the bullet.
They were then prepped and loaded as described above, and fed to Bertha:
I'm sure I'm not the first person to run this test, but I wanted to show decisively the effect poor bullet casting has on target results. YMMV.