@gregorme created an half automatic way to download the XML file of a complete course, see here:

The trickiest (looking) part is the line that parses all relevant course page from the start page XML

sed "s/^.*\[\[\($COURSE_PREFIX.*\)|.*\]\].*$/\1/" $ROOT_XML | grep "$COURSE_PREFIX"

If someone is working on the Export page (https://de.wikitolearn.org/Spezial:Exportieren), this could be used to more automate the process so that one only has to give the root page name and all subpages are selected (or added to the list) automatically.