We humans may know that, but do the algorithms know the difference between a video describing one voting error and a video describing multiple cases of errors?
It's not a rhetorical point because if you upload a YouTube video with controversial keywords, a banner appears below the video with clarification which may or may not be relevant. I once watched a Rammstein music video and was scratching my head as to why a banner appeared about the Apollo Space Program, as if Rammstein was a conspiracy theory and not a fictional music video. If this same algorithm is being used to censor, a lot of innocent videos will be deleted.
On the spectrum of "anecdotes about how a few of my friends didn't get their votes counted" to "fraud that swung the results of the election" it's clearly something close to the latter, not the former. And the latter has been repeatedly debunked.
I think you're willfully misreading in order to score some rhetorical point, but just in case:
Read that as "widespread (fraud or errors)". No one is alleging that the election was error-free, only that it was free from widespread error.