This is a major issue - and one that is misunderstood by so many. Wayne Tripp was very good to explain to me the limitations of retail AFR sensors. And he really made sure I understood the route I was looking at. I had no choice - as there is no alternative for me.
The simple fact, as Scott points out, is that the calibration of the "retail grade" wideband sensor WILL go off over time (I assume "use" rather than time). This can be tested on the PC5/AT - but I don't know if it can on the DoBeck.
THIS COULD RESULT in the O2 sensors actually measuring wrongly. i.e any display would "suggest" to you, that you are running at (say) 14.1 - where as you're actually at 13.8 or 14.5 - if analysed using a DYNO GRADE O2 sensor - which I'm told are 250-300USD a pop. Dyno grade sensors also "go off" but a good dyno-god will know this and only use correctly calibrated ones.
It's why I analyse (attempt to understand given I know what has changed) and then commit (and occasionally manually adjust) trims on the PC5. I know that I can simply remove the A/T box and have a "bespoke" base tune in the PC5. In fact this is my aim. I just seem to keep changing things so need to adjust. But the minute my AT trims are negligible against the committed trims map in the PC5 - off it comes.
My ideal situation would be to pass my PC5 data to a pure TuneECU hex map. Invisible. If I can get my hands on the right statistical software (I know what I want, as I used to sell it, but it's expensive and not crackable) - I will.
Is there a STATISTICIAN IN THE HOUSE?