I'm guessing the tweets referenced in that article were based off of the original[1] (that I saw). The original tweet series did a bunch of testing with different variations of the same image to attempt to get to the bottom of why the algorithm is doing what its doing. There was also an official response from Twitter[2].

[1]: https://twitter.com/bascule/status/1307440596668182528 [2]: https://twitter.com/twittercomms/status/1307739940424359936?...