[n4r9]

The only explicit experimental setup I can find on that webpage is in the last couple of paragraphs of Q.36, where it describes a "reversible machine intelligence" which can perform measurements and then reverse the entire measurement process. The issue is that this reversal requires having fine control over the joint quantum state of the system, measurement apparatus and all record of the measurement outcome. However, the very fact that this whole setup can be described as being in one quantum state or another is sufficient for a Copenhagen advocate to deduce that no collapse would have occurred.

Now, different flavours of Copenhagen will differ as to why collapse has not occured. Bohr would probably have argued that a measurement entails an interaction between a classical system and a quantum one, whereas this has not happened in the above. A more sophisticated approach like QBism would say that quantum theory is simply an agent's calculus of experience, and that if the entire system has a quantum representation then this means the agent is still external to the system and has not yet interacted with it. Either way, Copenhagen makes the same predictions as MWI in this experiment, as it does in all experiments that admit a purely quantum description.