So decrease your kerf width in the tool by 50% of the error. If you primary goal is the inside dimension, you 0.50" hole is measuring 0.480. That means you need the torch closer to the line by 1/2 of the .02 inches. Decrease the kerf width in tool by .01 inch.
That is why I suggested that you pick the dimension that is most important to you. For me, when I am making brackets, it is generally the inside dimension of the holes and the hole to hole dimension. The outside dimension of the bracket usually has more tolerances for my purposes.
If you are wanting a compromise, and lets face it, it is all compromises. You will never have them both perfect. I feel I have my numbers dialed in but even at that, I always add a minimum of 0.02 to 0.03 inches for bolt clearances. I would split the difference and go for a kerf of 0.055 inches. You are not talking about wasting a lot of metal in either case.
And as Knick pointed out, there could easily be a slight bevel that could be fooling you. And as Phillip and John point out: Your numbers are good and probably better than many.
The numbers in the first post are worst case at the base of the bevel, away from the cut surface. I"m envisioning what WOULD / could get in the way of a slot/slot interface (dinosaur skele puzzle recently is driving this… WAYYYYY too much hand filing to fit).I care far more about internal features than external.
I’m no expert here but with my 82i all the parameters you had set up I assumed you were cutting 1/4", not 10ga. Even with 1/4" you’d use a 1.0 nozzle, not a 1.1
That said, I have my 10ga dialed in at 40amps, 1.0 nozzle, .8 pierce, 85ipm and 65-70 psi and I think I get great results with that.
I don’t even bother putting on a 1.1 until I’m over 1/4" thickness.
I just cut two full sheets of 3/16" using your parameters you listed but with a 1.0 nozzle and 65ipm and it cruised right through it