First test with the small launch engine SEEM to indicate that the results CAN be useful.
At least, once I managed to figure out how to get a halfway decent reading with such tiny amounts of testable gas (about 1/6th of a cu. ft. per minute is not much), I also immediately learned that a change in setting, with this measuring principle, takes TIME to effect a change in readout.
I also learned that this engine too, needs several minutes to stabilize its behaviour after a change in settings.

The engine is externally governed, which means that when changing the needle setting, the governor immediately counteracts, so the RPM stays the same, the exhaust note stays the same, and tuning by ear becomes totally impossible. The only tell tale signs basically is the stability of the governor output signal (how much the throttle servo has to correct to keep the engine on set RPM).

Long story short, it took me about 10 minutes to get a reading I could thrust. It turned out that CO level was very high at 10% before adjustment.
It took 5~6 minutes of carefully manouvering with the needle to get the CO level down to 2,5~3%, and once that was accomplished, it took another 3~4 minutes before the new setting took effect and the throttle servo gradually became less nervous and active.
Exhaust effluent also became near invisible and a lot less smelly.

Going to let the engine cool down completely, then do a 2nd run on the same settings.
I expect a bit of a stubborn engine as long as it is cold, because I had the needle leaned out about 7 or 8 clicks, leaner than I can remember it ever was...

What can I say? It SEEMS to be a viable way of adjusting model engines, but I still have to figure out a way to get good exhaust gas sampling on a plane engine (I expect the propwash to be a practical difficulty to overcome).