When you test the jets, do you measure them for flow or measure the hole?
A 200 jet has nominal diameter 0.048" and a 210 has 0.049", so one could measure to that accuracy if forced to, although flow depends as the square of diameter so measurement errors are magnified. Unfortunately, the flow also depends linearly on the "contraction coefficient" of the edge of the jet, which varies from 0.62 for a sharp edge to 0.97 for a rounded edge. From this it can be seen that slight differences in the shape have significant effects on the flow. I'm told operators sat at the jet calibrating instrument (a flow meter with fixed pressure head, held at constant temperature 15C and with standard fuel of density 0.710 g/cu.cm) and slightly burnished the edge of each jet until it flowed what it should.
As historical speculation, there was close cooperation between Amal and Mikuni and Mikuni jets
almost have the same numbers as Amal. The slope of flow rate vs. jet number for Mikuni jets isn't quite as steep so my speculation is they used the same testing rig as Amal did, but the density of their "calibration fuel" was a tiny bit lower and/or the viscosity was a tiny bit higher.
Fluids, of which air is one, have similar flow behavior if they are in the same range of Reynolds numbers. Somewhere I have my calculation that shows that light air being rapidly sucked through a jet on my flow bench is in the same range as much denser gasoline flowing being sucked more slowly through the carburetor. This makes life easy.
I have a plate that bolts to my flow bench that has a single tapped hole for a jet. I also have a jet that I soldered shut that I use to 'zero' the unit. It's not like I jet bikes for a living so I can "calibrate" a set of larger and smaller ones than I think I might need so when I find the bike I'm working on a little too lean at 200 I can replace it with a 210 knowing that it's a 210.
I should say I rely on an average curve of cfm I measured on the flow bench vs. jet number, not on an absolute calibration. Thanks to eBay I have a huge collection of jets from 50 to 1700. Taking the 500s I mentioned earlier, 8 of the 12 clustered closely together in air flow with the other 3 outliers. For my purposes I ignore those 3 and draw my curve through the center of the cluster. Having done this for other jets I have a curve I can rely on when checking any other random jet.