Thanks for this excellent overview of the current available models.
I'd only note that when you say a model is `terrible`, all of this image generation technology is mind-blowing compared to what you could expect to do on consumer grade hardware just a few years ago. We've come a very long way in a very short time.
That's absolutely true. I got into this well before Stable Diffusion existed, so I have galleries full of pictures from VQGAN et al, and SD 1.5 was already amazing by comparison.
