Small sample set aside, the performance differences here are much bigger than I've seen in previous linux comparisons. Something has to be off right? Curious if anyone is able to reproduce these results.
Yeah, 'y' every time should be pretty safe for updates, right? Like, the reason it's suggested is because it's the suggested alternative for the package it's replacing and the change has been tested by the CachyOS/Arch team(s) already.
And worst case, something big breaks when you reboot, so you roll back to before the update, wait like a week, and try again (lol).