I think all those are good suggestions. I have heard of O2 sensors getting lazy and degrading MPG before a CEL is invoked, or just taking longer to go from open to closed loop - wasting gas in the process but not forcing a light.
It just is really tough to tell - this is one of the reasons why it is important to log MPG every tank - you know when something is not right. To most, the time and hassle of this is not worth gallons of fuel wasted, so they just wait until there is a sign.
Im not so sure that using adds in the right doses, low doses, is going to do anything to the engine. If there are loads of inorganics, from, say, running a non TC-W3 2-cycle engine oil in there, one may see an issue... but most adds arent all that different than the parent fuel they are dosed into, and they are in such low concentrations, I doubt there is the potential for an issue.
Doesnt mean that burning oil, excessive doping of adds, and all the rest won't destroy the sensors.
Is there a reason to think that the sensor needs to be changed? any signs of degraded performance? If you have one, you might want to put a scangauge on and see how quickly you go into closed loop... Without an "as new" reference, its tough to compare, but looking at at least that, and preferably the waveform is the only real way to know...