It's not whether it "really" looks similar. It's what people think, most of the people, and most of the people are neither known for practising good writing nor consuming good writing.

I use dash a lot while people rather usually use and are used to seeing a hyphen. I was called out on a certain app "wtf dude.. the least u can do is nt use ai". Well, the person was using shorthand and textpeak a lot, so it was already getting nauseating for me, so this outburst helped me eject, but not before I politely asked why they thought so and dash was the trigger along with "all da time crct grmr and spelling". Also "hu da hell writes dis long sentences". Guilty as charged.

Apple's proofread is essentially spell-check and punctuation until it isn't and even in a few-sentence-long para you'd see it has sneakily changed a lot and Apple being Apple you, the customer, obviously has no way to set it to "only fix spelling, punctuations and leave everything else including grammar as it is" and I've a feeling a lot of folks are at least using proofread or something on those lines. But then I really don't think browser's "spell check" ought to be kosher either if the content has to be the human's because those mistakes are also makes such text human and in some way unique. I don't think it's an easy line to draw but weird seeing just comments "targeted" here.

I'd rather ask AI to provide a source and then cite the source. But if the source itself is AI backed, then it's a bit different :)

I wasn't sure whether it was an omission or an unintended gap, as the guideline specifically points to "comments". So it seems AI generated/edited posts are fine. Strange, because both can be flagged/downvoted if it was to be left with that.