I dn, all the pictures of my friends and family and dog look pretty close to reality. I'm not sure I care if a cloud is slightly different, to be honest.
OP probably does more than it seems by interpreting what their client is asking for. Clients ask for some weird shit sometimes, and being able to parse the nonsense and get to the meat is where a lot of skill comes into play.
I think Cleo Abrams on YT recently tackled this exact question. She tried to generate art using DALL-E along with a professional artist, and after letting the public vote blindly, the pro artist clearly 'made' better content, even though they were both just typing into a text prompt.
