David Chalmers talks about 'fading' or 'dancing' qualia and predicts that one possibility will be that some kind of 'dancing' or 'fading' qualia will occur as the neurons are replaced. He also argues that any such 'dancing' or 'fading' qualia is very unlikely. The Idealized Effing Theory World
provides one idealized theory making precise predictions of just what kind of 'fading' and 'dancing' qualia will occur, and what it will subjectively be like, as you replace various neurons and systems in the brain, and why.
This idealized theory illustrates how the typical description of the transmigration process completely leaves out any system capable of 'binding' our conscious knowledge painted with phenomenal properties, so we can know, absolutely, what redness is like, and how this is different than greenness. Our assumption is that leaving out this binding process is the 'fallacy' leading one to use mistaken assumptions like a phenomenal property can 'arise' from anything that is functioning properly, wither it be neurons or silicon. Our prediction is that Chalmers' "Principal of Organization Invariance" is leading people astray in their effort to understand and discover phenomenal properties, and the fundamental qualities of consciousness.
Within the Idealized Effing Theory World
there is the all important binding neuron that unifies the palate of neurotransmitters that a brain is using to represent our phenomenal conscious knowledge of a strawberry with. Obviously, if you replace any one voxel neuron and the particular neurotransmitter it is firing to paint a red point in our conscious knowledge of the strawberry, the prediction is that the binding neuron, or process, will easily know that nothing but glutamate has the right quality. In other words, there won't be any kind of 'fading', you simply won't be able to fool this binding system with any single simulated version of glutamate, on any single location representing the strawberry with redness.
The prediction is also that you will be able to replace the entire binding neuron, and all the neurons 'painting' our phenomenally colored knowledge of the strawberry by firing with the appropriate neurotransmitter, in such a way that whatever is modeling the entire system can be interpreted in such a way that it will enable it to 'pass' the Turing test, or claim that the simulated glutamate has a redness phenomenal property. But, since we will effigy know otherwise, we will know what why and how the zombie system using only abstracted knowledge is lying.
Our assumption is that reality will turn out to be much more like this idealized world, and that Chalmers is wrong when he argues that 'fading' or 'dancing' qualia are not very likely to turn out to be true.