It takes practice, just like anything else. And understanding limits.
Love the Mads Mikkelsen.
Whether deBoar is right or not, you made a clear point how can art made with AI flourish on its own in the future, similar to how photography fought for itself and broke away from imitating paintings.
Sorry if I'm misrepresenting something, as I did not read the deBoar article, but his complaints are pretty insignificant. There many reasons why any given model will not give you poor approximations of a celebrity. But you can fix it. You can roll up your sleeves and train your own LORA, or better yet, see if somebody already has, here is one for John Candy: https://civitai.com/models/77388. Sadly, I couldn't find one for Goldie Hawn.
You are right about B&W, and about men.
Think what you will be doing next year.
Looking forward to seeing your art.
As to whether or not it will change the world, well, who can say? I have the same basic issue with futurists and Luddites: We can't really predict what's coming or the many ways that technology will impact our lives and culture. We can make predictions but these are often wrong or too limited in scope, and there's often good with the bad.