Thanks for the input, I appreciate it.
I have not tested that specific Hugging Face / FLUX approach yet, but I may give it a try as an experiment.
My main concern is still production reliability: licensing, API access, consistent quality, cost, long-term availability, and whether it can actually fit my fusion workflow.
So as a test, yes, definitely interesting. As a real production replacement for Fusiomon, it sounds at least difficult/troubling.