Ok, I didn't do a full ten, but I did test 5-6, and they all came out at about 1.040-1.045.
Using your "subtract .015" method, I loaded a small batch at 1.030 (at 5% less than minimum published charge), and they all pass the plunk test. Almost all rotate after the test too, so this seems to be about where I'm at with this bullet combo.
I tried calling XTreme awhile ago about some 7.62x39 plated bullets and velocity data, and it was very difficult to get anything from them. They finally gave me a number of 1500fps, and maybe I could get away with faster velocities, they had heard of others doing so.
I didn't exactly hang up with a warm-fuzzy feeling, or any great data either.
Hopefully it was the person I called, and not indicative of the whole customer-service experience there.
To be honest, I'm not really that particular about how I "feel" about a given OAL, other than some heresay I've read about the shorter OAL. I am however very wary of loading outside of published data, in this case 0.05 outside of the data, which is where I start to feel nervous about the whole deal.
Hopefully my 1.030 loads will yield clean cycling, plus give me some indication that the pressure is still fine. I can work up from there, slowly I guess, unless I can get data from the manufacturer as you mention (I prefer this, at least as a starting point).
I think it's the fact I've only been reloading for a couple of years; if I'd been doing this for 20 I probably would have a much better "feel" for the whole margin-of-error factor. Right now I mostly have what I can read, in reloading manuals and online, and most of it tells me there is very little margin in 9mm, given the case size (which makes perfect sense, from a physics angle).
Thanks again for all your help, hopefully this gives more to go on,
-pete