Second paragraph is a response to @latchkey's question

  > How do we know their bots didn't have bugs in them that pushed the results in favor of the tone of the article?

And I'm replying, that the point of the report is to tell about information bubble. The same one you are describing.

  > how they watch videos and how they scroll through them, and as such changing recommendation algorithm

Bots' viewing rules can only change the *amount of war related videos* uploaded before March 4th, not the *reported lack of fresh videos* from the other side for the russian bot. And as such there should be no issues about the lack of source code.