right, OK so when you say "a cat is classical" you essentially mean that due to decoherence, quantum effects (like diffraction of the cat) are much more difficult. and the interference between alive and dead states of the cat becomes very small. The problem is that still, the interference is non-zero. And the Copenhagen interpretation 'solution' to this problem is that we impose a subjective collapse which makes the interference between alive and dead states exactly equal to zero. This gives us a way to say "what probability is the cat alive?" (in other words, we impose the subjective collapse so that we can give meaning to the probabilities of certain states).That's exactly what I am NOT saying. I believe everything is quantum and in principle you can demonstrate quantum effects at all levels
Now this 'solution' raises the further question: "how small does the interference between alive and dead states need to be, when we impose the subjective collapse?" the answer of course, is that it just needs to be small enough to be within the experimental error of our specific experiment. In other words, we must make sure that any consequence of the subjective collapse is not detectable in the given experiment. There is nothing fundamental about this choice. The system has unitary evolution as far as we know from experiment, therefore we must impose this subjective (non-unitary) collapse at a time where it will not cause the predictions to change by more than the experimental error. So the Copenhagen interpretation 'works', but it is definitely not what we would hope for in a physical theory.