Required Accuracy for Filter Offsets

With the nearly full moon I thought it was a good time to check my narrowband filter offsets. I was imaging with a scale of around 1900 with my AT8RC and moonlite focuser. Steps size is set at 50. Filters are Baader.

The largest offset was for SII which averaged about -24. The Ha and OIII filters were usually in the single digits but occasionally in the lower teens. These changes represent .5 of a step for the SII down to between .02 to .28 of a step for the other 2 filters. As they were not completely consistent and the variance is so small, I’m tempted to ignore the variable small offsets and treat the Ha and OIII as parfocal. What do you guys think?