Thank you for your answers.
I would, however, appreciate some more details.
There is a small possibility certain hardware malfunctions become affected by a power cycle, but that is entirely unique to the specific failure condition and likely NOT your issue.
What makes you think that it's not my case?
I hear Windows can do weird things to your BIOS/UEFI settings.
I assume any changes to BIOS should remain even after turning off a computer, shouldn't they? That would imply that I need to boot into Window only once and then all would be fine.
The only difference is your finger pressing the on/off switch.
If the only difference between restart and switching off/on a computer were pressing the on/off button, how would you explain this empirical observation:
1) turning on and booting into Debian => some hardware doesn't work properly ( in my particular example my MIC )
2) turning on and booting into Windows with immediate restart into Debian => hardware works fine,
When hardware works fine in Debian, restart laptop and boot into Debian => hardware still works fine
Switching off from a working Debian session with a subsequent boot directly into Debian => hardware doesn't work
3) turning on and booting into Windows, switching off, switching on with booting into Debian => hardware doesn't work
I am sorry to bother you with a probably stupid question, but my only reasonable explanation for this observation is that there must indeed be some difference.
My naive hypothesis is:
Windows drivers set my snd card inner state register into proper state and this state is preserved during reboot ( probably volatile memory ) but not during switch off/on cycle. Debian then better handles my snd card in this working state than in a default state set, probably by BIOS, during cold start.