I still don't know what your point is. don't use llms to remove bias as they also have bias? is it just nihilism for the sake of nihilism? If so I can kind of get that, but then it leads to nowhere right? if something doesn't work perfectly I'd still use it if it's better than a less good alternative. I see it as a matter of relativism.

Your claim is that LLMs are better at "x" than the alternatives, for some value of x.

My point is that your claim is unsubstantiated.