To detect a photon—to know that it was there—you have to destroy it. Seeing absorbs photons. Yet you've just seen that what's between source and detector does affect the photons, so there must be some connection.
Hard calculations connect emission, absorbing and what's in-between. Here is a way of reasoning that mimics the calculations, without having to do them.
Each photon explores all the space between its emission and detection. As it explores that space, so the arrow spins.
The longer the photon spends exploring a path, the greater the trip time
, so the more the arrow has spun. The photon explores all paths, and each path contributes. The photon does not go down any one path: it explores
the lot.
Each explored path contributes an arrow. Total all these contributions to predict the brightness—that is how many photons the detector will absorb every second.
It might seem crazy, but that's the way the small-scale world seems to work. If you want your predictions to work, do the calculations like this.