Hm, true, interesting. Didn't notice that (only) the 90µm UltraT setting was changed somewhere between 1.4.x and 1.6.x since I never typically use anything below 140µm.
Is that intentional? It seems weird that Z-ABS >=90µm uses 0.5mm and Z-UltraT >=140µm uses 0.5mm but only Z-UltraT at 90µm uses three times the retraction.
Just checked: in v1.5 the setting for 90µm was still 0.5mm, so the change to 1.5mm happened when the way of storing the profiles was changed for the 1st time.
Is there a chance that something went wrong when converting the profiles from XML to source code???