You may be able to simulate a magnet perfectly, but that doesn't make magnetism. In the same way, you may be able to simulate a mind perfectly, but that doesn't make consciousness.
With consciousness, it's about experiencing sensations, what things feel like from the inside, not about any behaviour on the outside.
I think you've exposed the dis-analogy by making this argument.
A magnet is the cause/source/reason for/generator of/<insert verb> of a magnetic field. If you put a test charge in a field you can detect a magnet.
So you argue, no simulation of a magnet produces a magnetic field. I assume you mean that you can produce data which would characterize a field were the real world in the state that you simulated.
Then you say a simulated mind doesn't produce consciousness because consciousness is internal not external. Simulating a mind and interacting with it is just behavior checks.
This doesn't make sense then. With magnets you say, "there needs to be a physical field present." But with minds you say, "there needs to be undetectable internal experiences, just reporting and observing behavior isn't consciousness"
This analogy just begs the question. The issue of consciousness is whether it is purely a functional/dynamic state of a system or whether it is a basic atomic kind that cannot be further decomposed. For example, a simulation of carbon isn't carbon but a simulation of disorder is in fact disordered in some sense.
