Eric mentioned in linked article, that huge number of joules is not needed.
Your math isn’t wrong, but I was referring to surge strips (not necessarily a strip, but not a whole house unit). We don’t need joules for series mode protection. It’s a parallel mode thing which I don’t recommend in a strip.
For in-panel protectors, which are parallel (and the only one’s available), they publish surge current instead of joules but the effect is the same. Repeated surge current wears those MOVs down.
Of course, surge protectors are a lot like air bags. No guarantee you’ll survive, no guarantee they’ll ever even be needed but still the math says cars with air bags are statistically safer than cars without them.
That does NOTHING to protect you from the start-up spike put on the line on a single 110V phase when you turn on a vacuum cleaner, table saw, refrigerator or other significant motor load anywhere inside the perimeter protection of a WHSP unit.
I’m not sure this is completely true. As I understand it, the limiting factors are the clamping voltage of the WHSP, as well as the inductance in the line in between the surge source and the WHSP.
It IS true that the best surge strips (again not necessarily a strip, could be a rackable device) like Furman and Tripp Lite have lower clamping voltage and can reduce the surge effect at the TV, for instance, especially if the vacuum and the TV are on the same line.
PS - Given that all panel mounted surge protectors have about the same clamping voltage, I prefer the ones that mount as breakers. This minimizes the wiring and I hope minimizes the impedance, helping the surge protector to lower the surge voltage. Plus I think they are easier to install/remove. Right now I’m using a Siemens BoltShield, but previously had a Square-D panel and did the same.