[AllyCat]

....  the precision of the reading depends mainly of the precision of R4 and R5 resistors, as also the precision of the internal reference of the ADC, which you can't control.

Yes, even with exactly the "correct" resistor values, the value of the "Reference" voltage may produce a significant error.  But I believe the reference is actually the "external" (5 volt Vcc) rail, not the Atmel's internal regulator.

If the Transmitter is "on the roof", it is likely that so will be the battery and the calibration multimeter?   IMHO it is more logical to "tweak" a value in the Transmitter software than to change resistors (or to fit in a potentiometer).  This also covers the use of more than one Transmitter (with different calibration necessities) and avoids the need to edit every version of the Receiver software (which is generally updated more frequently than the Transmitter software).

Obviously the value could be "calibrated" within the transmitter software and the Receiver software left unchanged., but currently there isn't a transmitter "calculation" in which to "tweak" the value.  Also, I was trying to reduce the software size in the (original Pro) Receiver, because I want to modify the Fan (aspiration) control, which may make the program too large (but that's for another thread).  Similarly, I've been considering modifications to the Transmitter power supply arrangements (e.g. using Lithium Cells) where a higher measurement accuracy may be required and with perhaps the Nano's "5 volt" supply rail no longer accurately regulated at 5 volts.

Cheers,  Alan.