Now that I’m more aware of the problem, I’ll try to pay closer attention as to exactly when it is occurring. In general I do quite a lot of manual filament calibrations, and I now wonder whether the actual calibration process is contributing to the problem, since the first pass in Orca slicer’s flow rate calibration will generally lead to both under-extrusion and over-extrusion during printing of the calibration tiles. Again, I’ll try to pay more attention now to observe whether or not this is the case now that I’m more aware of the problem.
As for filament drying, I 100% agree and I’m already all-in on that. I even purchased a blast oven to do it properly, together with using desiccant to keep it dry afterward.
I know some nozzle vendors advertise that they use nickle coated nozzles because they claim it affords a less sticky surface for hot extruded plastic to cling on to. Unfortunately, AFAIK, Bambulab doesn’t offer any nozzles with either that or similar no-stick/low-stick finishes.