Thanks for that!
The Delay Choice thing makes sense to me; I followed that okay. The Eraser Experiment seems completely lost on me, though. Either it seems to be logically flawed or I'm simply not 'getting' it (I suspect the latter). However, it seems to rest on some kind of magical 50/50 probability issue that I'm not too comfortable with, as it's not very precise about what causes it and why.
Let me try to put some of thoughts into better words here, as this is still swimming around my (already tired) head this morning...
Regarding the Delay Choice experiment: A photon is sent towards two slits (A and B), with the possibility that it will either have a wave formation (and therefore travel through both slits) or a particle formation (and therefore travel through only one slit [A or B]). I presume there must be something that determines whether it is A or B, but I guess it is largely irrelevant for this discussion. The
real question is what causes it to be either a wave or a particle, and that's what we will find out at the end.
So, the photon has travelled through a slit, or both slits. We don't know yet, because we can't detect it at the moment. In order for it to have passed through both slits, it needs to be a wave pattern. In order for it to pass through only one slit, it needs to be a particle pattern. So now comes the part where we detect that...
We set up a screen at the far end which can detect the light hitting it. All experiments show a wave-like pattern which proves that the light must have travelled through both slits in the form of a wave. Every experiment that uses a screen to detect the photons will give this same result.
Now comes the clever part: It is possible to very quickly remove the screen. So quickly, in fact, that it can be removed faster than the light can reach it
after it's already passed through the slits (in order for this to be reasonably possible, the screen would have to be a very long way from the slits -- but it is physically possible to do).
Behind the screen are two telescopic devices each trained onto one of the slits. When looking through these devices you can see a flash of light to indicate that the photon has travelled through the slit that the telescope is looking at. It will either be one or the other... but it is
never both.
Now, this is interesting, because if the light doesn't appear in both devices, how can it be a wave pattern? It can't; it has to be a particle pattern and therefore has only travelled through one slit.
So, to clarify: When we observe the results on the screen, we see that the light travelled through both slits. When we observe the results through the telescope, we see the light has travelled through only one slit.
Since we can make that choice
after the light has already passed the slits, we would assume that the answer already exists and we just don't know it yet, but it would appear that this isn't the case, because it would otherwise be a complete coincidence every time. What we discover is that our choice whether or not to move the screen or not move the screen is what
causes a result that has already happened. Essentially, an action in the present affects something that happened in the past.
How bizarre!

Have I understood it properly?
If so, here are some problems and possibilities:
1. We're assuming that the photon cannot be both a wave
and a particle. Is it possible that it is both? Will the screen detect the photon as a particle or is it only designed to detect it as a wave?
Let's imagine (and feel free to laugh at my probably totally incorrect physics, here) that the light travels from the source in both a wave
and particle pattern. Putting a screen there will detect the wave distribution pattern but won't detect the particle pattern (which is there also but isn't detectable).
Not having the screen there will show the particle pattern but not the wave pattern (which is still there but isn't detectable). I don't know how exactly these things are detected, but it seems strange to me that the sreen will
only detect the wave and the telescopes will
only detect the particles.
Even if (and I'm assuming it can) the telescopes
could detect the wave (ie. light passes through both slits and into both telescopes) we couldn't say that we chose it unless we know for sure that the particles
never hit the screen. Do we know that, or do we simply know that they aren't
detectable on the screen? Or
are they detectable on the screen but it just never happens?
A little help here, please!!
2. We're assuming (perhaps -- again, feel free to laugh uncontrollably at my total lack of physics understanding) that the photon 'travel' in a traditional sense in a forwards motion. What if it doesn't? What if photons 'travel' by being everywhere at once, and will only be at a later position if it has already been decided (the instant it leaves) that it can reach the place it's going to? We know that time dilation would make the passage of time to any photon instantaneous and infinite. ie. It would know, in
its time, whether or not the screen is there before it has left to choose a slit by
our time reference. This is getting tricky for me now... -- What if the reason the photon is detected one way or the other depending on our method, is because the photon knows that method before we've done it, because by
its own time frame, it has already happened. Wouldn't it be similar to the reason why electricity won't travel through a circuit unless that circuit is complete? Or a similar reason at least?
Or have I inadvertantly explained the exact same result as the experiment originally suggested albeit from the perspective of the photon instead?
Agh, my head hurts!
Right, I'll tackle the Eraser thing later. I need to lie down.
