nice documentation and it looks as you are very close to the optimum.
you could try to lower the z level adjustment value approximately 30 - 50 µm more.
there is an "optimum" when doing these tests. if your nozzle is to close to the printbed, the thickness of the sheet rises up again (because the surface becomes more rough and the squeezed edges of each strand tend to overlap).
but take in mind: when you've found the optimal z-level adjustment value, the adhesion to the printbed is maximal. which might cause big prints with large bottom surfaces to stick at the printbed as hell
so it might be a good idea to alter the z-value to a slightly higher level than the optimum... otherwise you might get problems to get your print off from the pei sheet.
dem inscheniör is' nix zu schwör...