Originally posted by photoptimist LOL!
The only way to get "medium format quality" is to have "medium format sensor size." Basic optical and photosensor physics strictly limits image quality and affects the "look" of the image.
The computational image manipulations used by smartphones are like pancake makeup on a corpse -- sure, things look pretty, smooth, and lifelike but the true image is rotten underneath.
If the end result is pleasing does anyone outside of photography forums care if the data to construct it was collected with one exposure from a big sensor, or many exposures from a smaller one? I'm having difficulty discerning the relative purity of: light hitting sensor > processor collecting data > processor turning data into 1s and 0s in some format > storing data on memory card, and light hitting sensor > processor collecting data > processor turning data into 1s and 0s in some format > processor optimizing and combining multiple exposures > storing data on memory card.
It's splitting of hairs to rave about pixel shift doing computational tricks on a K-1ii, and then talk of smartphones combining exposures as garish makeup on a corpse.