After producing a large number of my aluminum filament drives, I've been looking at my "reject pile":
when i make one of these things, i do things a little differently, every single one is done by hand on a small American built lathe , the aim is to produce a filament drive with both consistant roundness in the hobbed area and sharp well defined teeth , the actual diameter of the hobbed section can end up anywhere between 10mm and 11mm this is almost irrelavant since you have to go through the extruder steps/per mm calibration procedure anyway
looking at my reject pile, the main reason for rejection is that the hobbing goes too deep or the gear goes out of roundness. if it goes out of roundness they end up in the bin, the rest are generally get given away or used for testing, the irritation is that for every ten good ones i normally get at least 3 rejects
just due to the hobbing going smaller than 10.50mm however they are all still round to less than 0.05mm and realistically make little to no difference on most direct drive extruders anyway