It's different because we know for certain that the mapping of words such as "good" and "puppy" onto our basic phenomenology is culturally dependent, learned throughout our childhood, etc. We can say no such thing about the mapping between physics and subjective experience.

I'm missing something here. How does the fact that this correlation isn't as culturally dependent imply something spooky is going on?

It's not because we can't precisely map them; it's because the possiblity of there even being a mapping is so weird and complicated that spooky, exotic physics looks good by comparison.

Again, I don't follow your logic. What would be weird and complicated about such a mapping?

Basically you would be forced to argue that the mapping between classical physics and subjective perceptions was picked by an optimizing agent, which is far more spooky.)

Why? What need is there for an optimizing agent? What do you think this optimizing agent would have done? I'm not sure what you are trying to say here but it almost seems to some sort of argument that if one wants to reject theism one needs spooky physics. I don't know how to respond to that.

We can see macroscopic correlates of emotions and sensations in an fMRI. Ths does not mean that the emotion and sensation is the same thing as the change in fMRI. (In fact, all fMRI does is measure changes in blood flow.)

You might notice that I said "fMRI and other methods." We can for example, use deep brain stimulation to directly stimulate emotions (this is in fact a cutting edge treatment for people with severe depression and is being investigated for use in treating other illnesses). We can see which parts of the brain are being used for what emotions and sensations and we can stimulate those regions to duplicate those emotions and sensations.

More generally, it seems like you may be confusing the map with the territory. A blank or poorly drawn area of a map doesn't tell us about the territory. It is true that repeated failure to get a good map of an area of territory can tell us that our mapping method has a problem or that another section of our map has issues. That's essentially what happened with Copernicus and Kepler; the repeated failures to get accurate models of the heavens forced a redrawing of fundamental sections of the map. But in order to justify that, one needs to have repeated problems over a long period of time with trying to get a good map of an area. If your map keeps getting more and more precise, that's not useful. Finally, a question if you don't mind: what hypothetical evidence would convince you that qualia can be explained by our current laws of physics?

We can say no such thing about the mapping between physics and subjective experience.

The wavelength of light maps pretty straightforwardly onto our perception of color. We can trace the activation of cones in our eyes to patterns of neuron firing in the optic nerve to neurons firing in the visual cortex. "Redness" isn't magic. "Redness" is a particular configuration (or, more properly, a set of configurations) of neurons. The only reason it seems special to you is because you are experiencing the algorithm from the inside. Consciousness is what thinking feels like, not magic.