Some sites hide content as it loads in, using some combo of display or visibility CSS properties and JS to detect window load.
Generally speaking, this is a bad practice as you can greatly delay the perception of speed on your page.
While this behavior is pretty obvious when you see it (a blank page for several seconds and then suddenly the entirely rendered site appears at once), I was wondering if WPT metrics could provide a signal for this.
I was thinking about using FCP vs. Start Render. Since FCP is a browser metric and Start Render is a lab metric, a discrepancy between when the browser signals loaded visual content and when content actually loads it could be indicative of content being hidden.
Would that be a valid approach, or does FCP not get fooled by this hiding technique?