Lawrence, Jason. The 'delayed choice' simply resolves on choosing different starting assumptions. Yes, Jason, as in QED's sum over paths and Huygens construction (foundational to quantum and laser optics and photonics) there can be no 'photon' or path' until the combined Schrodinger sphere surfaces (NLS equation) are forced to interact with matter (a 'detector') where only ONE position has adequate constructive interference to quantize the new particle (at different 'ranges' the 'positions' also differ).
'Non-locality' then emerges from particle 'reversibility', which 'weak (statistical) measurement' can't discern. I invoke electron 'spin flip', which in fact Bell also did, relating measurement to direction of DETECTOR field electron spin - which REVERSES with EM field orientation (setting angle)! Bell also admitted his (Bohr) assumption; Bertlmann's sock always differ. A sphere has BOTH spins (poles) and a sock can be randomly worn inside out (pink becomes green). OAM of a sphere is conserved through x any y axis rotation. 'Direction' is NOT conserved!!!!! That's the DFM derivation of 'non-locality'.
To explain in terms of Wheelers view and (i.e. Jacques) delayed choice;
The focussed waves follow BOTH 'paths' from splitter 1, so each detector has a 50:50 chance of clicking.
Introducing a second splitter COMBINES them, so phase can be tuned so EITHER detector can have 100% constructive interference, leaving the other 0%.
As Wheeler said; "No elementary phenomenon is a phenomenon until it is a registered (observed) phenomenon". It is wrong to speak of the "route" of the photon in the experiment of the beam splitter. It is wrong to attribute a tangibility to the photon in all its travel from the point of entry to its last instant of flight."
Timed pair experiments then CAN access data which gives A,B aa or bb if just one detector dial is reversed. My previous essay identifies that 99.999% of Aspects data confirmed this, but couldn't theoretically rationalise it so discarded it. Wieghs (et al inc. Zeilinger) found the same so just 'corrected' for it! Perfect examples of 'theory bias' in experimentation.
My short summary completes the work in the link above by classically explaining non-locality as well as entanglement.
Classical reproduction of quantum correlations. But we seem now permanently trapped by 'theory bias'. Can you now see the solution Lawrence? Most surely won't.
Jason
"The good news is that once we figure this out, it will most likely lead to new physics. Maybe we'll find out that we have grey alien neighbors." The figuring out was the simple bit. It's done. The real job seems to be is to overcome our human failings to make it visible!.