The 2nd method you describe is not how a CO sensor works. Unfortunately is far more complicated than a resistance plot. It works on a difference of voltage in the control circuit that runs it. Subbing in a simple resistor wont bypass it. The control circuit can operate as a simple on/off switch based on the sensors output such as killing the ignition spark on the kill wire terminal. That control circuit may be far more complex such as outputting a pulse signal to the inverter module, to which that module translates as an on/off state for the inverter output and kills the ignition. Without that pulse signal, the inverter will shut down the generator after a few seconds. Sadly these sensors are far from simple in many cases.
Take this sensor for example:
https://www.digikey.com/en/products/detail/spec-sensors-llc/110-102/6136363
All CO sensors have a shelf and service life. That is all determined by RH and temperature. It also responds differently too based on those specs. So how you store your generator, if the RH/Temp is frequently high, it shortens the life expectancy.
Typically they are rated for 10 years of service life, but at ideal environmental conditions. More realistically in practice, these sensors will see about 5-7years before failing. My technician contact at Firman told me that the control circuit design they had on their sensors were too oversensitive, so a "bad" batch made the sensors problematic. They have since corrected it, and if you do have a faulty sensor, it is easily replaced without digging into the wiring/voiding warranty.
The circuit design in general is intriguing to me, but I can see how complexity causes diagnostic difficulty. As someone else posted, Sensors cannot fix stupid!