Take the measurement at the nozzle with or without a windbreak.
At a point in the near future you can tell by a puff of smoke or just the force of the breeze
I need about 40cfh minimum with my typical contact tip/ nozzle combination in still air, and continue to get good results to 50cfh, which works better outdoors for me. Much beyond 50cfh and I begin to get unpredictable results, but numbers are different for different tip and nozzle variations.
The issue is VELOCITY of the shield gas at the nozzle tip, not really volume determined by cfh at the flowmeter. Nozzle and tip variations affect the velocity as if it were a garden hose carrying water it would affect the spray pattern. The reality of the dynamics is only slightly more complex.
With excessive velocity at the tip [flow rate too high] you risk something like a venturi effect from turbulence drawing air into the shield... you should know where this point is with your welder; experiment with high flow rates in still air and know how far you cn go when in a breeze. Extra volume seems to work better in a breeze. This does not seem to be an exact science and I've never seen rock solid specifics cast in stone as for transition currents, etc.
Flow rates are perhaps the most forgiving of GMAW variables. When forced to I throw economy to the wind and crank open the valves... sorry for the pun but I hope it will help a bit.
I recall a safety issue of the AWS Journal and dimly remember guidelines about ventilation and rates of air flow for ventilation. It may have been a different publication for a different application but I am reasonably sure.
BTW the nickel in stainless qualifies as toxic and apparently causes lung cancer.
This edit reflects a correction of misimpressions I previously held, awareness of which I owe to our very able and professional system administrators who, by the way, deserve a bit of recognition for their ability to keep this system up, dependable, and civil despite an array of challenges.
Regards
d