New classical system to produce correlations ?

[Dirkman]

This looks interesting . Can someone please take a look at this ?

http://vixra.org/pdf/1609.0129v1.pdf

[Schmelzer]

The beginning is already bad:

It is almost universally supposed, that “Quantum Correlations”, as discussed in connection with Bell’s Inequality Theorem, cannot be produced by any classical, macroscopic system. 

No, it is not.  To obtain Bell correlations with classical macroscopic systems is easy, even trivial.  The point is that every such system needs some information transfer, which, in the conditions where Bell theorem violations become interesting, have to be superluminal, thus, forbidden by relativity.  

I was unable to understand exactly the meaning of the different pictures, but it seems that 

In any real experiment, the detectors will not be able to detect the existence of every particle or coin-image.
...
Consequently, if a threshold is set, such that the absolute value of the correlation, as measured by the matched filters, must exceed the threshold, for both images in an entangled pair, then the Detection-Efficiency will drop below 100%. 

suggests that the effect is simply a variant of the well-known known detection efficiency loophole.