To keep my personal image catalog compact, I am considering scaling down all images that are unsharp for one reason or another, but are interesting or valuable enough not to delete them altogether. I am finding surprisingly many such images:
Some are simply out of focus, to varying degrees.
Some were taken with a compact camera, producing 12 megapixel images from a tiny sensor (a clear victim of the megapixel race). Viewed at even 50%, many of those photos look somewhat blotchy.
Some lost detail or acquired severe artifacts from repeated JPEG processing in my less quality-aware days, but are still large files.
The candidates include processed DNG files. Given the lack of sharpness, these are far from my best works, so I am willing to give up the advantages of the raw format there.
I am wondering if I would lose any useful information if I scaled all those photos down to a quarter of the pixel count (or even less; where possible keeping an integer ratio in each dimension for cleanest scaling). Could future improvements in processing technologies make me regret that choice, or do these images really just take up an unnecessary amount of storage space?