In the 2010s, PageSpeed Insights' metrics compared to your site to how it could be if it was delivered optimally. Did you compress your images? Have you combin

In the 2010s, PageSpeed Insights' metrics compared to your site to how it could be if it was delivered optimally. Did you compress your images? Have you combined your CSS files? Are your static resources longcached? A huge site perfectly delivered would score high, while a tiny site sloppily delivered would score low. They represented how much room for improvement your site had if you didn't want to make deep changes. Search was doing its own thing, and they didn't ever have any public speed metrics. Core Web Vitals is a different approach: pages are compared to thresholds. A huge site perfectly delivered now likely scores lower than a tiny site sloppily delivered. (PageSpeed Insights now reports CWV metrics.) These are the kind of metrics you can use for search ranking, because they are about user experience instead of developer room for improvement (low-hanging fruit). (I used to work on mod_pagespeed, which would automatically optimize sites, and I still work for Google. Speaking only for myself.)