[renormalised (Carl)]

Nothing too onerous about that experiment. All they've done is measure the average momentum of a packet of photons by passing 50/50 polarised light through a calcite crystal. Because the 50/50 polarisation will cause the photons to enter the calcite at slightly different angles, the refractive properties of the calcite will then skew the paths of the photons slightly. This will mean that more photons will end up polarised in a certain direction than the other. But measuring each individual photon would give the expected result of a 50/50 chance of polarisation direction. That beam was then split and the average momentum and positions of the photon packets determined by the intensity of the two different interference patterns i.e which direction of polarisation was made up of more photons (in essence, which pattern was brighter). They could then trace the average path of the photons by repeating their experiment and moving the screen they projected them onto, creating a map of the average path. Tedious, but they could measure both the average momentum and positions of the particles "simultaneously".