PageSpeed results are not perfectly static. Network conditions, simulated device performance and other test conditions can move scores slightly from one run to the next.
SiteVitals keeps the history so those normal fluctuations can be seen in context. When a meaningful regression appears, you have the previous results to compare against rather than a single isolated score with nothing behind it.
Because performance lives on the same SiteVitals timeline as uptime incidents, content changes and other site events, you can also see what else changed around the same time. That makes it much easier to investigate whether a slowdown followed a release, an image upload, a third-party change or something else entirely.