Hi @Raphael Huber,
First of all, I apologise for the late reply! I must have overlooked this note.
Thank you for the interesting question. After some thoughts, I thought one way to achieve what you want in Pandas is to create intermediate/temporary Boolean columns. I have prepared a starter code that explains this at the following GitHub gist:
If you run the code, you will still have those intemediate columns. Then, exporting the styled DataFrame as an Excel file and removing the unnecessary columns would be a fairly straightforward step.
Of course, your use-case may be more complex than this, but I hope this starter code helps!