But "95%" is an arbitrary number used because an early statistics book had a table of 5% confindence widths. (Things were done by hand then.) The "correct" confidence level is not a mathematical concept but is set by the practical consequences of making one decision or another.
As a practical ansatz, yes. But if something still has a 5% chance of being true, you haven't falsified it.
Secondly, if we can never tell that any other O-regions exist, then the entire concept of O-regions is "non-falsifiable" empirically.
And that gives me paws pause it what purports to be *science* which is based on systematic observation, experiment, yada yada. You know the drill.
Cheers!