Agreed in practical terms that this "try it and see" method is the only realistic way for the majority to make any progress; AND if such progress and improvement is even possible.
The only alternatives that I am aware of, is to let the router auto-switch and/or directional aerials and/or relative positioning of router and WiFi PC.
The best WiFi channel may vary on an almost continuous basis, remembering the effects that pulley/weight driven tall clocks could have on the early TV transmissions around 50 MHz and 180 MHz.
Unfortunately, most WiFi users are unaware of all the other devices using the 2.5 GHz Band; and the WiFi monitors present such as inSSIDer only show the detected WiFi signals in a very cleaned-up manner, as though there is nothing else using that part of the spectrum.
Then the radiated power levels of WiFi are extremely low; and the assembies very small - one I have to hand is 5 cm long by 3 cm wide by 4 mm (milli-metres) deep. A lot of that is screening and the basic PCB, so the active components must be almost microscopic.