I pretty much assume that "minimum thickness" applies to a structural wall, not details. Rejection could easily be that a machine operator doesn't understand, or is over-enforcing their preflight checklist. That extends to preflighting software, too - my direct experience is preflight software frequently flags issues that are no problem in execution. OTOH, if you blow-off the preflight checks, you're out the time and materials cost when something doesn't work.
I agree on the Z-axis thing. Several of the equipment manufacturers spec their X/Y in ppi and then the build table in microns. (Which triggers a pet peeve - mixing spec units to obfuscate performance. Like in the Pinshape example, taking the "best" spec as the published resolution. Not.)