I think this is highly unlikely to be a targeted attempt to censor Mastodon. The simpler explanation (and the explanation with most historical evidence behind it) is just that Facebook's AI algorithms are kind of bad and nobody in the company understands how they work or what associations they build.

However, the underlying idea that Facebook would block links to competitors is historically valid. As recently as 2016, Facebook blocked links to competing networks from Instagram (https://www.theverge.com/2016/3/3/11157124/instagram-blocks-...), and leaked internal emails from Facebook have shown that the company has an extremely broad view of what does and doesn't count as a competitor (https://panatimes.com/facebook-bought-instagram-to-neutraliz...). The company is extremely anti-competitive, it's not shy about this, and internal emails show that this anti-competitive attitude is entrenched very deeply and very consciously within upper management.

I think taking down this post in specific is very unlikely to be deliberate because:

A) Mastodon is likely not a large enough service to warrant it, and because

B) The explanation based on Facebook's AI being weird, opaque, and generally untested is a much cleaner, simpler explanation that requires fewer jumps in logic.

But it would be completely in character for Facebook to target a real competitor in this way. The reason it's unlikely to be deliberate is not because Facebook would never do something like this, and it's not because Facebook would be too frightened of regulators to do it so openly. Facebook has very openly done stuff like this in the past. It's just that there are other explanations that are more likely, and it's that if Facebook was going to start doing this, Mastodon probably wouldn't be among the first competitors they would target. I need a lot more evidence to show that this is deliberate before I jump off of the (extremely compelling) explanation that automated moderation is really buggy across the board and regularly does unexpected things.

The article comes off as a little uncurious to me, I feel like the author is jumping too quickly to a specific conclusion without a lot of critical thought. But part of why Facebook has these problems with people jumping to conclusions about how it tracks and moderates is because Facebook has a very real history of being openly corrupt in these areas, and Facebook has a real history of being deceptive about their motivations behind decision-making processes. The reputation hasn't come out of nowhere.