Yeah but the bots that people complain about are the malicious bots that try to pass as humans.
I think the only problem would be if this API was used by non-bots (perhaps some relays) and the tag loses its meaning.
Discord does something similar, although I feel they've gone way too far with policing it.
What they don't realise is that the comment they are reading right now could well have been written by a bot.