Interesting. I had come to very similar conclusions on the 22lr vs 308. I was running mine on a rather smaller data set and I came up with 25%. I had my first Gen targets made but I'm working up new ones that will give more feedback.
What I think is critical with the 22lr is that there is more feedback on misses. I think for 22lr training, you should be able to see where every shot fell- exactly. Hits and misses are good, but knowing where you missed is almost more important. That's why I don't like shooting 22lr practice or competitions and get no call on misses.
That feedback needs to be TIGHT to get the most benefit out of it.
https://www.snipershide.com/shootin...f-7-62x51-175gr-out-of-a-18inch-r700.6395085/