HI, if you have decently sized dataset (large variety of artist images), it would be pretty easy to create LoRA for this style. For style LoRA it's better to have 100+ images at least.
If we'll talk about prompt, i'm sure it would be almost impossbile to replecate this style because it's too complicated, as i think.
I've seen some old yiffy e18 models doing something similar, but it still was a lot simpler.
maybe try this prompt "furry style:1.2, body fur:1.2, wolf face:1, (legs fur, arms fur), furry body:1.1,[:fur:0.8], woter color:0.8, [water color|sketch], sketch:1"