Note that for the screen printed samples, I changed the peak definition.
For the samples doctor bladed on glass, I defined a peak as >20 micron.
Many of the printed samples (on TPU) didn't have those high peaks.
Therefore, I changed the peak definition to >10 micron.
This should not change much for the baseline thickness 
(potentially, the reported baselines are a bit higher for the >20 micron peaks, as a valley (i.e. baseline point) is defined as a valley located below the average peak height (in order to exclude shoulder peaks).