Read that Chesterton quote somebody posted above. It sounds like you need to study the standard a bit more, because you don’t seem to understand why NaNs are the way they are. You’re arguing from principles that don’t apply.
I don't know enough to say; all I remember from the few interactions I've had with SQL over my long programming career is that it's intellectually unsavory as a whole.
