The GFM 2.0 doesn't "sniff" a leak rate. It does two simple things at once: it measures the total flow passing through it (using a venturi - a smooth narrowing in the pipe; gas speeds up in the throat and its pressure drops by an amount that reveals the flow), and it measures what fraction of that flow is methane (the sensor). Multiply the two:
Leak rate (CFM) = total flow (CFM) × methane fraction
That multiplication is only honest if every bit of the leaking gas passes through the meter. Miss 20 % of the gas at the cone and your answer is 20 % low - and nothing on the screen will warn you. The entire rest of this page exists to answer one question: will all the gas actually make it to the meter at your pole length?
Diluting the leak with ambient air is fine - the math above still works (more flow, lower concentration, same product). Losing gas is what ruins a measurement.