Sounds like a plan.
And as you suggest it would not be difficult to arrange a practical test, my idea would be;
At a location instead of a Gateway use a simple test transmitter that sends out test packets at the appropriate power and spreading factor.
Then take a matching simple receiver and in a few directions and travelling away from the TX check at the actual locations where reception ends.
Now use a planning tool to model the location and see what model parameters most closely match the reality.
I appreciate that people do the modelling and check RSSI readings by looking at the gateway log, but we know that the RSSI reported by the LoRa device is at best an ‘estimate’ that has changed substantially since LoRa came out, so I would not be assuming its in any way an accurate indication of the real signal level.