1. They had someone with a clear history of strong anti-Trump and anti-Republican sentiment take the action (https://twitter.com/LevineJonathan/status/126545757821512499...)
2. Twitter chose a prediction rather than a factual statement to fact check ("Mail-In Ballots will be..."). Why not start with a truly factually wrong statement about the past?
3. They picked something that is actually debatable! A bipartisan committee concluded it carried some risks in 2005: https://www.wsj.com/articles/heed-jimmy-carter-on-the-danger...
The notion that a company can ever be trusted to "fact check" (aka determine objective truth) is just completely laughable. The closest we can come is labeling agent beliefs about truth ("X says Y is false").
Doing nothing would be better than doing this. Even better would be building solutions that allow community-based (and ideally personalized) derivations of consensus (this is what we're doing at LBRY).