I'm thinking that light slows down in a medium because photons are being absorbed and then remitted by atoms or molecules in the beam. This would imply that the photons which leave a lens or filter are not the ones which entered. How would this impact the an experiment designed to demonstrate entanglement?
First of, photons do not have identity. It is meaningless to say things like "photons which leave a lens or filter are not the ones which entered". More to the point of your question though: if a photon that is entangled with another system gets absorbed by an atom, that atom is then entangled with that other system. When an atom that is entangled with another system emits a photon, that photon is entangled with that other system. In this chain of events entanglement gets passed on.
I'm thinking that light slows down in a medium because photons are being absorbed and then remitted by atoms or molecules in the beam.
That is one possible process and it takes place for example with glass. It happens, what oleg told in his answer: Doing an experiment with polarized light, the glass doesn’t destroy the polarization. The light is coming out with a delay but still polarized. So your conclusion
This would imply that the photons which leave a lens or filter are not the ones which entered.
does not apply in all cases.
Your question touches another interesting point. As we know an edge influences photons. Your question about re-emitted photons is one of three scenarios which occurs:
- The photons are re-emitted from the obstacle.
- The photons propagate undisturbed in some distance from the edge.
- The photons interact with the edge by their electric and magnetic fields and the directions these field components get changed.
Polarizers are used to rotate the electric or magnetic field of light in certain directions. To prove this influence of the polarizers, you place three polarizers in series, each polarizer being rotated by 45° to the previous one. The light goes through this arrangement.
The interesting thing comes now: If you remove the polarizer in the middle, no more light can leave the last polarizer.
The real question is, is the rotation a quantized process and is the accompanying deflection of the photon also quantized? A positive answer would mean that the fringes behind the edges are the result of quantized interactions between the edges and the photons.
Even in one-to-one photon experiments, the edges appear on the screen, even for a single edge! The standard explanation at the moment is given only for single slits: the wave-behavior of the photon interacts with both edges of the slit and the fringes are an interference pattern. For me, an explanation with quantized deflections is more intuitive.