Intervening as if there were a recession inminent when it is not also has harms (the exact same as the harms when recession interventions are maintained too long or employed too intensely, in terms of inflation, etc.), so I wouldn't agree that your central bank is bad if you happened to have guessed right once, but only if you have a demonstrably accurate objective method.
It actually appears that the alternative harms aren't as bad; that is, recessions aren't caused by prior expansions and we don't get one by "deserving" them.
Is there a way to predict when central banks aren't "good enough" then?
Point is, for a 100% positive case rate, I have to tolerate a 22% false positive rate. What exactly is the complaint here? Was this line of logic meant to make people who make these predictions look stupid or foolish somehow? To me, it mostly fails to.