RAW is supposed to be the data that the sensor captures (well, today that's more or less true). The actual images saved on phone/SD (JPEG or HEIC or whatever) a

RAW is supposed to be the data that the sensor captures (well, today that's more or less true). The actual images saved on phone/SD (JPEG or HEIC or whatever) are just processed RAWs with some destructive presets (WB, contrast curve, sharpening, color tone-mapping, ...) applied and written out to storage. Some of these presets are destructive (gamma curves, bit-depth reduction for JPEG) and cannot be undone. So of course RAW doesn't make for better photos, but for original input data to do your own edit on. If you don't know how to edit, you're probably better off with the experience/work that the manufacturer put into the in-camera presets. In my opinion the main advantage is that RAW files save the full color-depth info _and_ RAW files are not gamma-mapped (i.e. linear encoded), that enables all of your use-cases above. You would still get 90% of that advantage if you were to demosaic, color correct to a profile-connection-space but keep the linear encoding and original 12-16 bit depth, even if you were to compress that data via DCT/JPEG to save space. That's sort of what ProRAW is.