Measurements of spin correlations: Alice goes back to Bob and compares results

Question

Take a fully entangled two-photon state, like

$|\Psi>=\frac{1}{\sqrt{2}}\big(|+->-|-+>\big)=\frac{i}{\sqrt{2}}\big( |yx> - |xy> \big)$

where one photon is sent to Bob and the other to Alice. This is the typical situation for the EPR paradox using spin states. Thanks to the No-Communication theorem we know that Alice cannot send any information to Bob, since, whatever angle she chooses for her spin detector, Bob will still measure random outcomes. Only when the two get back together, can they observe that there is perfect correlation between their measurements. All clear up to here, very well repeated in hundreds textbooks.

Now the question: I am interested in a simple experiment that proves what written in all those textbooks, namely, do you know any measurement where Bob comes back to Alice and compares the results? I do not mean any Bell's inequality test, which is far less intuitive, or any super-uper complicated experiment maybe difficult to follow. I just want a simple measurement (possibly old, since nowadays nobody would be interested in this) where, given the same source of maximally entangled photon pairs, two detectors (Bob and Alice) measure:

first bunch of photons: Alice detects along circular base (+-), Bob detects along circular base (+-)

second bunch of photons: Alice detects along linear base (xy), Bob detects along linear base (xy)

third bunch of photons: Alice detects along circular base (+-), Bob detects along linear base (xy)

then compare the results and see that we have perfect correlation for the first 2 bunches, while Bob measured uncorrelated stuff in the third bunch. Or something like that.

Answer 1

I discovered that what I was looking for is called 'coincidence measurements' of entangled photons.

Many experiments measure the photon coincidence in the linear base, like in Fig. 2(a) of 'Ultra-bright source of polarization-entangled photons' from 'Kwiat, et al.', available at arXiv:9810003.

On the other hand, fewer measurements are performed in the linear and circular base at the same time with the same source, like in 'Two-photon entanglement in type-II parametric down-conversion', PRA 50, 23, Figs. 3,4 and 5. These experiments reproduce the first two scenarios I listed in the question. The result for the third scenario can be inferred, noting that a single circularly polarized photon would give no preferred measurement direction in the linear basis.

These experiments prove what written in textbooks, i.e. that entangled photons give full correlation when Alice and Bob come back together, if they measure in the same base, whatever that is.