Required tweaks to Doxygen exports

[ChristophWeber]

To be imported into the developer portal, we need the Doxygen XHTML datasets to be reworked a tiny bit. As posted at her at the time of this writing, all four datasets need one or two additional tweaks.

1. All files should be in the base folder once unpacked. I.e. no html/ subfolder. I worked on this for QA purposes and used zip -jj systemcoupling.zip ./systemcoupling/* to get the desired result. A working example dataset is at https://github.com/ChristophWeber/DevRelDocs/raw/main/veronica.zip
2. It would help if we agree on a standard file extension. The veronica set uses .xhtml, while the others use .html. Our initial code works with .xhtml only. You can control this with HTML_FILE_EXTENSION in Doxygen. We will have code shortly that can deal with both, but it is still better to have one agreed on process that is well tested. In particular, extensions should never be mixed in the same dataset.
   Thanks for your attention to this!