I seem to recall a video (perhaps KurzGesagt) that a magnitude 25 earthquake would overcome the binding energy of Earth - the planet would be in separate pieces. That definitely is an upper limit: when the "earth" in earthquake literally cannot take more.
Earthquakes are sort of naturally limited though. A 9.0 is going to be catastrophic no matter what, and while I’m not saying a 10 couldn’t happen it would probably be something like once in a billion year event.
10.0 is firmly in "if it happens nobody's gonna be around to care what it's designated" territory I think. There's a practical point at which the death tolls are going to be sufficiently high that the number probably shouldn't matter.
Though in tornadoes there definitely are EF-4 designated twisters that are hotly contested online as being truly EF-5; often that's down to where damage occurs in the lifetime of a tornado though and it being difficult to prove windspeeds when a system is moving through, e.g.: trailer park vs an industrial park.
I don't think it's quite as cut and dried as that. A 9.5 hit Chile about 60 years ago, and about 95% of the most directly hit town survived. Which is not to minimize it - there were thousands of fatalities - but it was human scale tragedy, not apocalypse.
It’s actually a function of energy released. The earthquake can get bigger if the fault slip is larger. A magnitude 12 quake is technically possible but requires an entire hemisphere to slip 500 meters. There’s a really interesting paper that takes the moment magnitude scale to its logical extremes.