> English is a bastard of a language and getting messier every day as new nations adopt it is their standard language.
I disagree strenuously with this idea, because it suggests that there is one 'big' English in which anything goes. A better idea is the one of the register[0]: there are many Englishes, many sets of rules. Different rules are used in different regions, by different groups of people, and have different connotations (e.g., the King James Bible was intentionally written in a form of English that was considered archaic at the time because that would make it sound more grandiose).
If I were to use this tool, I'd be using it to ensure that whatever I'm writing is well-received by my intended audience. Because English usage is so varied, I would need to be able to control the register that it uses to ensure that the output is suitable. The fact that the product website doesn't even mention a list of supported languages, let alone supported dialects and registers within those languages, has a very everyone can see what a horse is kind of feeling[1].
[0] https://en.wikipedia.org/wiki/Register_(sociolinguistics)
> there are many Englishes, many sets of rules.
Absolutely, but try to make a run-of-the-mill LLM understand this.