I also had some test loads in 223 to compare neck-turned brass with non-neck-turned. I have not measured the actual dimensions yet, but through the scope I did not see a significant difference in accuracy, but the velocity spreads were half with the neck turned. That should translate to more accuracy at longer distances. It was an eye opener for me.
Ok, had a chance to look at the numbers.
Loads were 22.7, 22.9, 23.1, and 23.3gn. Previously, 23.1 was my accuracy load.
Non-turned necks
22.7gn, 2661 fps, SD=13.9, ES=26.24.
22.9gn, 2689 fps, SD=16.0, ES=38.2.
23.1gn, 2712 fps, SD=8.4, ES=21.22.
23.3gn, 2754 fps, SD=15.1, ES=35.3.
Neck turned:
22.7gn, 2678 fps, SD=6.5, ES=18.4.
22.9gn, 2700 fps, SD=7.8, ES=19.8.
23.1gn, 2733 fps, SD=9.5, ES=23.4.
23.3gn, 2749 fps, SD=7.1, ES=18.3.
Interesting to note the consistency across the charge range with the neck-turned loads. The only one that was not better was the 23.1gn load, and the difference there was not statistically significant. I can only assume that the more consistent velocities were due to more consistent neck tension.
All of the groups were 1.0" - 1.4", with the groups showing some horizontal spread that I would associate with the 10-15mph crosswind that was present, so I can't really say that any one group was any better than the others. The 23.3gn loads both showed the least amount of vertical spread, though, at about 0.5".