Wow, lots on stators. Don't want to argue, just providing my experience.
On unit construction bikes, if I have trouble getting .008", I change/fix whatever is wrong. I've never had a stator/rotor rub except when a main bearing went - can be either side.
Except Norton I say away from non-unit construction bikes - just not my interest area.
For Norton I work to get .010 at the tightest spot. There are so many things flexing in a Norton that I like a little insurance. When I have trouble:
1) I start by measuring the stator spacers. I've found them to be three different lengths - up to about .005" between the longest and shortest. Moving them around sometimes resolves the issue. A way to tell this is a problem without taking anything apart is: find the smallest feeler gauge that will go all the way in at the tightest spot and then the largest that will go only part way in - if different, the stator is cocked.
2) I check the inside of the stator. In some cases, the encapsulation is extending inward past the metal. If I had the means to do it. I would use a mill or lathe to fix that and also take off a little. I like that better than reducing the rotor, especially if the rotor is reduced with high-speed steel inserts. Rubbing steel with steel and creating localized heat on magnets are both known to be bad for them. Nigle's efforts show it not to be a problem, so my concern is moot.
3) The stator should slip on the studs with minimal effort. If it doesn't, I use a tapered reamer to slightly open both sides of the stator holes but it does not touch most of the depth of the hole. If it still doesn't slip on, I change the studs. If it still doesn't, I try a different stator (I have a couple of original, but bad stators I use for this test). If that one doesn't slip on, I try to determine which stud is out of line and open the corresponding stator hole with a round file. If still no joy, I replace the inner primary.
You would think that the inner primary attached to the engine and to the cradle (only when the outer cover is on) could not flex in relation to the crank. I'm not so sure. I've found too many loose/improperly shimmed cradle to inner primary studs, loose engine to inner primary bolts, chains hitting the inner primary and so on.
Let's ignore chank flex. How about the built-in slop (C3) in the main bearings. On each revolution one piston is under great pressure downward and the other is not. There could easily be no runout manually turning the crank and quite a bit with the engine running. I can't find the C3 specification for the NJ306E.M1.C3 bearings most of us use but I'm guessing based on other bearings around 30 microns (.001"). So, one down and one up but the rotor is close to one (making my head hurt).
On any British bike other than Norton, I'm sure skimming the rotor doesn't do much to the output. I say this because the same rotor and stator work on Norton and they are not aligned so the magnet's effect on the stator are already different. Since it works for Nigel, it's probably OK there too. However, I get the impression that Nigel keeps the revs up (we all should). I bet if he installed an ammeter, still has the bridge rectifier and Zenor, turned on the lights, and road at 2500 rpm or less he would be running on battery. Even worse if running on a half-wave rectifier rather than a bridge. Yes, a "high output" stator would help, but what they really are is more turns of wire, so the magnets have more potential effect.
