If anything this post disagrees with the claim that ‘English isn't generic for language’. It considers a naïve approach that fails for English, and shows that it fails for Russian in a similar way, and then compares it with an improvement also developed for English, which too helps Russian in a similar way.

It is true that when Facebook trained their model “blindly”, this introduced meaningful bugs, so I agree with that claim from the post. It does not, however, give credence to the idea that techniques developed for English don't reliably work for other languages.

Unfortunately the specific comparisons they make are suspect given how different the training sets were.