Relation between full step length and layer height

[Nicolinux 288]

I am setting up my 3DR delta printer and while tinkering with Slic3r I came across this statement regarding z-wobble and how to counter it:

Undulations in the walls of a print may be due to wobble in the Z axis. A thorough analysis of the potential causes is given by whosawhatsis in his article "Taxonomy of Z axis artifacts in extrusion-based 3d printing", however one point of particular interest for users of Slic3r is the wobble caused by motor steps not matching the pitch of the Z rods thread. This can be addressed by ensuring the Layer Height setting is a multiple of the full step length.

Source:

• http://manual.slic3r.org/troubleshooting/troubleshooting#z-wobble
  

The question is - what is the full step length for an UM2 stepper?

 

[Dim3nsioneer 557]

200 steps for one stepper rotation divided by 8mm thread pitch... or rather the inverse of it: 0.04mm.

[nightwish 0]

If I understand right this means we should try printing with layer height 0.08 or 0.12 as these are multiples of 0.04?

 

[3dmaker4u 0]

My question is, if we cannot rely on microstepping on Z, why could we on X and Y.

I accept that, in general, dimensions in X and Y are larger, hence the error is less significant.

But the point is that everybody is calculating resolution taken into account microstepping as well.

So, the 5 microns (perhaps 1/8 microstepping?) that UM2 specification claims on Z are reliable or not?