Not really, you can't. For example, you might have a service that lies dormant using 0.1mb of memory and then does some scanning/maintenance once a week and use

Not really, you can't. For example, you might have a service that lies dormant using 0.1mb of memory and then does some scanning/maintenance once a week and uses 50% CPU for 5 minues with 100m memory. Also, checking those dozens of services will take a lot of time Unsuspecting users can't even measure space savings correctly since many components can be hardlinked to something else and deleting them will "free up" the space shown in the folder (the size standard size reporting on Windows is deeply inadequate), but not free up anything in reality. Or for some uwp apps you can't even see the size at all as the folders have a very dumb permission set, and these scripts remove those preinstalled uwp apps. That's the point of public benchmarking - so that ever single user doesn't have to waste time doing the mistaken comparisons on his own, so your suggestion doesn't help at all even in theory.