I'm trying to optimize the perceived quality of a particular website.
Using Google PageSpeed Insights, my site's homepage is given a (low) score of 27. Using GTmetrix, the same page is given a very decent PageSpeed score of 88.
It was my understanding that both metrics were the same. Are they not supposed to be? If not, what is the source for the GTmetrix version of PageSpeed?
This blog post on GTM's website throws important light on your question - https://gtmetrix.com/blog/the-difference-between-gtmetrix-pagespeed-insights-pingdom-tools-and-webpagetest/
Key Takeaway - Don’t compare these tools equally and wonder why your scores differ from tool to tool…
While they’re all likely based on the original, open-sourced Google PageSpeed library, some key differences should be pointed out. After Google made general scoring algorithm/recommendation changes years ago, they have not updated the open-sourced library since. The PageSpeed Insights online tool as a whole seems to contain a new set of rules (which isn’t open-source).
Don’t compare these tools equally and wonder why your scores differ from tool to tool; they all use different recommendation sets.
This is why you’ll see differing rules from each service, and why we deemed it necessary to update our own PageSpeed ruleset. Read more from our changes to the recommendations.
Having said all that, I would recommend giving this brilliant post on Kinsta a read - https://kinsta.com/blog/google-pagespeed-insights/
In a nutshell, it states that scores do not matter. Speed does!