Hey Tom; typically asyncing or deferring JS will result in all parts of the page (UI) that are rendered by JS to be shown later, so in case of a JS-heavy site, the page might visually load slower, but then again JS-heavy sites are a usually a performance-buster anyway 😉
What is pretty weird in the wizard result overview, is the huge differences in total page size. As AsyncJS only adds async or defer flag to <script tags, it does not impact page size. My best guess is that the deferring/ asyncing makes the JS load later (which is the goal), but in the meantime the “hero video” in the carousel keeps loading more data, thereby negatively impacting the performance as measured by GTmetrix.
What I would suggest; run a manual test on webpagetest.org (or pagespeed insights) without and with AsyncJS and compare some of the more relevant KPI’s (First Contentfull Paint, Largest Contentfull paint, onLoad time, cumulative layout shift, …).
hope this helps,
frank
