how useful is focus bracketing actually ?
I have done some experiments on focus accuracy using MTF. You can get an idea about it in my MTF thread here, and the mention on a nerd use of AF tune also there.
With my Z7 II , I find that one AF-tune step corresponds to about 6 micro meters shift of the focal plane at the sensor for fast lenses. The might be also the smallest possible step for bracketing. On top of that come statistical focus mis-placements ranging 3-15 micro meters RMS at the sensor depending on the lens model and lens sample. For example with my AF-S 105mm f/1.4 on FTZ it is RMS = 3 and for the Z 24-120 f/4 it is about RMS = 10, which is OK for the lens speed.
The geometrical optics DoF at the sensor is: circle of confusion diameter times F-number : for example 10 micro-meter x 1.4 = 14 micro meter for the fast lens.