With img2img we use an existing image as input and we can easily:
improve the image quality
reduce pixelation
upscale
create variations
turn photos into paintings and vice versa
eliminate texts and watermarks

Link to the workflows from the video:
https://drive.google.com/drive/folder...
If there's trouble with the Latent Upscale node, try this one:
https://github.com/city96/SDLatentU...

Link to video on ComfyUI Manager:
   • ComfyUI 03  ComfyUI Manager Install  

Large collection of upscalers:
https://openmodeldb.info/