Yes, if they didn't recommend the video, if it could only be found through search, there wouldn't be any issue with it.
Though even with search, if the search algorithms are biased, and prioritize things in a way that could be argued to be recommendation, like personalized results based on what it thinks you'll like for the given keywords, you could say that's also a problem, which is why I mentioned Google Search.
To put it another way, you'd expect that the easiest content to find and discover, and thus be shown and recommended would also be the most commonly held opinion and current scientific consensus. But you'd also want representation for less commonly held opinions and scientific hypothesis and claims. Just that, if 1 scientist is pushing some wildly different idea, and 99 others are pushing the same idea, you'd expect that the idea pushed by 99 scientist would be the easiest to come around and that it be clear that it is weighted to 99%, while the other would be a little harder to come around and it be clear that it's weighted to 1%.
The issue is those platforms fail heavily to represent things proportionally as such. Often, the outlying wild claim will be recommended more heavily, and be promoted over others. Even without algorithms, the people pushing wild claims generally employ amplification tactics by spamming and posting multiple times over and over, etc. to falsely boost their representation.
So that's the problem I see here. Banning the more problematic instances of this is their poor solution to keep their platforms operating in a way that doesn't offer proper representation (because it makes more money), but avoid the extreme cases that could put them in hot water.