Frequency question with F75-LTD

It's my understanding that the coil and detector are tuned to one frequency. If that is true,
and noise/ EMI are not an issue then which of the 7 frequency has the best factory matched?

I've observed the transmit coil output with a pickup coil connected to an oscilloscope, and varied the operating frequency. The variation in amplitude (ie. strength' of TX) was less than 5% from best to worst. The signal was the most sinusoidal at the F3/F4 frequency, and F1/F7 were the most distorted.
Bear in mind:
a ) Observing distortion can be misleading - these higher-frequency components get largely filtered out anyway, inside the machine.
b ) 5% variation is nothing, amounting to less than 1% depth difference.
c ) I haven't considered the receive circuitry - that will almost certainly have frequency-dependant behaviour.
d ) This is just an observation of MY machine/coil. Yours may be different.

Common sense says it ought to be best at the mid-frequency. Maybe if you had the time, you could try an accurate airtest on a common coin, like a US 5c nickel or clad quarter dollar. I doubt the difference is measureable, less than 1/4 inch (6mm).

Thanks for the in depth response, this makes sense.
I've been defaulting to F3, guess I'll continue....

In a test-garden with real dirt........... there is negligible measurable differential...... between F1 - F7,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, under one condition; that being: EMI is a 'constant'...and minimal.