News Courtesy of Yoast.com:
The best site speed tools
When we’re analyzing the SEO of a website we always check the site speed. Obviously, site speed is different when checking it from different locations. This is just one reason why speed tools do not always provide the same results. That’s why we use all these tools when testing a site (and do not rely on just one):
Google PageSpeed Insights splits mobile and desktop. It has recently been updated to use more real-world data to give more usable results. Also, its suggestions to fix speed issues have been improved. Pingdom’s Website Speedtest Tools allows for multiple locations and GTmetrix combines several checks nicely. WebPageTest has a few main checks it grades in an easy to understand manner. Google Lighthouse is built into Chrome and was originally meant to assess Progressive Web Apps. It, however, gives great insights into the page speed and user experience of your mobile site, based on real-world tests. We would recommend you use all of these tools to check your site speed. Combined they give a complete overview of the site speed of your site.
Ryan’s Take 
If I had to pick only one of these tools to use it would be GTMetrix. I believe it does a nice job of balancing user experience against optimizations. It is comprised of two scoring metrics; PageSpeed and YSlow. The PageSpeed calculation differs from Google’s current PageSpeed tool. It is based on an older build and is more forgiving when it comes to optimization tweaks.
Google PageSpeed still provides very valuable data when it comes to optimization. However, as I’ve learned, it might not suit your visitors best to make sacrifices in order to achieve the best score. Render-blocking scripts and CSS minification might be better left alone. For me, FOUT (flash of unstyled text) and FOUC (flash of unstyled content) should absolutely be avoided. I’d rather have a few marks deducted on a score than compromise user experience. Google font swapping and content moving around during page load is both unsightly and distracting. In most cases, you may shave a little bit of time off but is it really worth it?
The one complaint I have with GTMetrix is how they calculate the YSLOW score. I don’t think a website should be penalized for not using a CDN. That’s just wrong. However, not using a CDN can lead to a large gap between the difference in PageSpeed and YSLOW score. Why the CDN hate? It probably has to do with my recent testing of Cloudflare’s free plan. It didn’t go so well. I found loading times to be worse as well as some errors occasionally popping up. In theory, it is a good idea to serve content based on where your visitor is located. However, most website owners only want to target a local or national area. Optimizing your website, server, and applications can achieve great results without the need for a CDN. Therefore, it is extremely frustrating to be penalized on YSLOW (I typically average around a 69% score).
Most of these tools will report the “Must-Do’s” in optimization. No sacrifice or compromising. Things like leverage browser caching, image optimization/compression, and GZIP compression are all simple but often neglected. Focusing on these little things and then deciding on minification and deferring rendering-blocking scripts afterward is the best way to prioritize tweaks and optimizations.
