I believe @EricJ is correct. Many of us have noticed after installing an engine monitor that the fuel pressure has a lot of wild variations in a M20J (maybe other models also). The Garmin EIS seems more sensitive to this than the JPI which may have better filtering. The analog OEM gauges don't seem to be able to react fast enough to display this variation. There are posts on vansairforce about this phenomena with the IO-360 in RVs. So, I think it's going to happen in any airplane with a Lycoming IO-360 and a tight cowl. I prevailed upon @kortopates to peruse Savvy data and he said it is common.
The fuel lines in the engine compartment are all fire sleeved and thus well insulated. However, the mechanical fuel pump is not insulated and picks up conducted heat from the engine and radiated heat from the engine compartment. This can cause vapor bubbles to form in the pump. You can easily see this if you direct a heat gun to the fuel pump with the engine off and the mixture in ICO so that there is no flow. As the pump heats up, the fuel pressure will rise. Directing the heat gun at other fuel components has no effect.
During cruise power operation, there is enough fuel flow to minimize vapor issues, and the RSA servo is designed to be insensitive to inlet fuel pressure variations and to separate fuel from vapor, so the vapor induced pressure fluctuations don't seem to affect engine operation. In my airplane, if I run the engine at around 1000 rpm on the ground for an extended time on a hot day the engine compartment will get hot enough to cause vapor and make the engine run rough. If I increase power to about 1500 rpm for a few seconds, it seems to clear it out.