Okay, I finally conducted my test of lube (Neo Lube) vs no lube.
For this test, I used my .300NM. Exact same load for each, one had no lube in the neck, the other was lubed with Neo Lube after scrubbing carbon from the necks.
This was my load:
- Berger 220 LRHT
- Norma brass, once fired
- CCI 250 magnum primers
- 83.0 grains N570
Each set had 25 rounds. I would shoot 5 rounds from the no lube set, followed by 5 rounds that were lubed, over the Magnetospeed. I took my time shooting to not let the barrel get too hot, conducting this test (50 rounds total) over ~3.5 hours.
Here are the results:
24 rounds (last shot didn't read over chrono) with the set that wasn't lubed:
Max velocity: 2945
Min velocity: 2894
Average: 2910
SD: 12.0
ES: 51
25 rounds with the set that was lubed with Neo Lube:
Max velocity: 2942
Min velocity: 2899
Average: 2924
SD: 10.5
ES: 43
Throughout most of the test, the Neo Lubed set was a clear winner. SD's were solidly single digit, being almost half of the non-lubed set in some cases. But as the test went on, the non-lubed data set started to "settle in" more on velocities, with a reducing SD. Meanwhile, the lubed set was getting more volatility in the MV, leading to an increase in SD's. I've attached a graph that plots the muzzle velocities and rolling SDs for each data set, its somewhat intriguing and I can't say I've made total sense of it yet.
View attachment 8475563
ETA: New cleaned up graph
In conclusion, the Neo Lube data set was better for MV stability, however I can't explain the phenomena that occurs on the last 10 shots, when there appears to be increased volatility. Especially since the other data set seems to settle in at that stage. Another thing to point out is that the Neo Lube set had consistently higher MV's by about 15 fps.
ETA: after more thought and reflection, the two data sets are so incredibly close that it's hard to draw any strong conclusions in regards to one being better than the other. I was monitoring the SDs after every 5 shots for each set, and there was initially a pronounced difference for the first 15 shots between the two sets. I think that introduced a bias in my perception, but after reflecting on this more it appears that the data sets are in actuality closer in performance than I initially concluded. The "phenomena" is within the statistical variance for each data set.