I wonder if we contact the old head of Evopidia he would send over the code for the wikipedia dump processor software. We could process requested dumps 'in house' as it were. EnWikipedia would be the most difficult or at least longest though our computer processors are 3-4x or more faster now than in 2009-13, I doubt the dump processor is multithreaded but maybe yes as I recall it ran like SETI@home on distributed basis. We might get a 2020 and beyond edition, I recall him being very cool and added wikivoyage among other things when I asked.
Is there interest in me contacting the Evopedia guy?
I would need someone or a group with full desktops or servers and a real internet connection so I need to poll for both user interest and active dump processors before I make contact.