Depending on aluminum alloy and anodizing color hardcoat plating thickness can change from almost nothing to .015". However, hard anodizing usually becomes brittle after .0035". So usually plating of color is targeted for a processing window of .001 - .0025". If the shaft is to be black sometimes the process window is pushed to .003" or greater to better saturate the color. I'm not sure how much Easton puts on, but I do know I have never gotten pure black saturation on 7075 with a plating thickness less than .0017". Probably the reason I always got a tighter fit with my old Nibbs using X7 black compared to X7 blue.
You must realize, however, that 60% of plating thickness penetrates the material and 40% is buildup. So a .002" thickness is only an approximate .001" in buildup. If measuring the I.D. of a non-blind bore the diameter will change .002" or less (.001" per side) from the non-plated dimensions.
To clarify, I have not measured an Easton aluminum shaft in probably 25 years or more. And at this time I do not have the sharp point micrometer or proper comparator to make such an I.D. bore measurement within 10ths. But, back all those years ago Easton's wall thickness was between the machinists guidelines for +/- on .XXX & .XXXX Usually falling much closer to the +/- of the .XXXX (.0005). So buildup in anodize thickness wasn't a real issue. It could have changed over the years, but I doubt it has changed much.
I would assume since Easton states the sizing in 64ths that leaves acceptable parameters within limitations of fractions rather than being limited by machining guidelines. So a 22/64" shaft could be .344" +/- .0075" and still easily fall within the size 22/64" diameter. I have a few dozen 23/64 shafts here and all run within those parameters on the minus side. Actually all run within the minus side of the machinists +/- .XXX parameters running .3555 - .3560 instead of .359. With twin .015" flat blade micrometer tips I get exactly .0145" on the 2314 wall thickness. I could probably figure the correct fine point calculation within 10ths mathematically if we really needed to. But that would just be minutia at this point, wouldn't you agree?
In your case .010" smaller on O.D. calculation does seem rather high. Are you using a micrometer to get that I.D. measurement or are you using calipers? To get anywhere close with calipers on this size diameter you have to adjust for the .010-.015" flat against the bore I.D., and you also have the adjust for the offset too. I find on a bore diameter of this size that can give you as little as .002" false reading and as much as .005" false reading depending on whose calipers you have. If you get .0025" or more of false reading then we are once again within the parameters of a 64th inch fraction.
Myself I'd probably go off the actual O.D. dimension since that is the one you are going to get the most accurate. Then subtract your .038" and see where that puts your I.D. From there figure out what your interference fit is. If it is only .001" then you might be able to just heat up the aluminum and press fit it on - depending on the length of your footing. If you still end up being .003" then you can probably remove 1.5 (3/2) from the carbon easily with a green or brown Scotchbrite. That would probably give you a better mechanical bond surface for your glue anyway!
