I think there's an extreme bias on KYM towards English language memes that have mainstream notability (e.g. being on the news; virality in the form of hitting the front page on major platforms; celebrity). It's also bottlenecked by the memetic literacy of the researchers who approve submissions, as well as their methodology of tracking spread via first appearance.
There was a good article about the issue on the Internet Histories journal:
Here's a great diagram by @cris.v.gutierrez, from an earlier thread (Anyone else surprised at how few memes are on know your meme? - #4 by cris.v.gutierrez):
I think the way that people talk about and use platforms more and more as tools (and less as places or spaces, cf. Colonizing Pepe: Internet Memes as Cyberplaces (2018)) belies a shift in their metaphysical intuitions about the Internet. Many of the earlier memes from forums and imageboards were memes by users from someplace, and inter-platform rivalries would sometimes lead to competition over which community really started a meme. People still seem to understand the Internet in this way, since most users still have a platform that's their main home base (and they consume other platforms indirectly via screenshots and reposts). But it's not nearly as strict a pattern as it used to be.
Nowadays it's not uncommon for people to repost screenshots of their own tweets onto other platforms (and people seem to do the same to r/4chan, posting something weird to 4chan and then sharing it as a screenshot to the subreddit as a "found object"). This makes sense both practically (it's easier to create content using simple web interfaces; a single-image tweet imposes a form that's amenable to memes), as well as aesthetically, since screenshots of content from specific platforms have particular connotations which are highly relevant to memetic content.