Jonathan, (please excuse the informality, but I'm not sure of proper etiquette here),
You stated

"Each time you replace a neuron with a mass of silicon gates you lose one copy of rich experience."

How do you know this?  You also say 

"Dennett's suggestion that behavior can be equated with experience is simply ridiculous."

That's not much of an argument. Exactly why is it ridiculous? Also, I'm not aware of where Dennett makes that equation.  What if behaviour is a necessary outcome of experience but isn't the whole story?  Finally, you say

"But the intelligence and behaviour of the robot you get with replacing all cells is exactly the same as that of the original brain.

So you would end up with a p-zombie, a creature/robot/agent that will swear up and down that it experiences the richness of a beautiful sunset, the heartbreak of a child caught in a war zone, the appreciation of nuance in a well-played Mozart piece. If we tell it that it doesn't have "experience", then it may respond "well, I certainly have something which matches all of the descriptions of experience in the literature.  If it is not experience, what is it?  Do you have it?"

*