Viktor,
I agree with you that references like Machinery's Handbook have some limitations in that they do not necessarily properly account for the potentially significant effects of tool materials, tool coatings, etc. However, what does that have to do with 1045 steel and 303 stainless steel having different machinability ratings for a given hardness level? Machinability depends on material microstructure, and the microstructure of a plain carbon steel like 1045 (hardness of 225-275 HB) is different from the microstructure of an austentic stainless steel like Type 303. One can obtain a plain carbon steel like SAE 1045, an austenitic stainless steel like Type 303, a copper alloy like C17000 (2% beryllium-copper), and a titanium alloy like Ti-6Al-4V all with a hardness range of HB 225-275, but they will exhibit substantial differences in machinability because of differences in elasticity and plasticity, strain hardening, fracture behavior, etc. Perhaps I misunderstood what you wrote, but machinability cannot be directly correlated with hardness, most simply because hardness is not an intrinsic material, but the result of elasticity and plasticity effects.