Some of them, using 3 detectors, allow to test predictions of polarization
On Saturday, two LIGO plus one Virgo gravitational wave detectors announced the discovery of four new black hole mergers – those from Summer 2017. Why it took 1.4 years isn't entirely clear to me. Their IDs, GW170729, GW170809, GW170818 and GW170823, reveal they arrived in July or August and the first, late July detection is the most massive and most distant merger ever detected.
These two black holes merged some 5 billion years ago – at the same cosmic time when the Solar System was created over here – and out of some 80 solar masses, a whopping 5 solar masses of energy were radiated away in the form of gravitational waves. The collaborations have released some new papers (see also the LIGO paper server).
It's really cool. Imagine: When the Sun and the Earth were being born, elsewhere in the equally cold and young Universe, some 80 times amount of mass than the Solar System did something else, namely merged two pre-existing black holes to create a bigger one. The event also sent some sound in the vacuum – which was thought not to exist according to Newton. That sound carried 5 times more energy than the Sun's \(E=mc^2\) so that despite the weakness of gravity and the billions of years it needed to travel and get diluted, some life forms on Earth could detect it.
One new paper is a catalog while the other paper extracts population properties.
Because I have mentioned Virgo, the Italian 3/4-smaller sister of its larger American brothers (at least I concluded that Virgo is a sister, not a brother), some of these mergers were seen with three detectors simultaneously. Well, GW170814 – the August 14th merger – was the first one seen by three detectors. LIGO and Virgo say that this allowed them to test predictions of general relativity concerning the polarization of the gravitational waves for the first time.
Three detectors are better than two but I think that the two detectors are enough to prove the mere existence of the two different polarized components of the wave. In fact, I think that the methodology of the crackpot papers claiming that no waves from the Cosmos were seen, if done correctly, could be formulated as a proof that the gravitational waves are composed of two unequivalent functions of time.
Why? Because these particular crackpots notice that if they subtract the "best possible signal", the residuals as seen in the two detectors are still correlated with each other. But that outcome is surely predicted if you assume that the two detectors detect two slightly different functions – polarizations along different L-shaped axis pairs. If the detectors hear similar but not identical chirps, it's clear that any approximation by a single non-stereo chirp will leave residuals that are still similar but not identical.
The real question is just whether there could also be a different plausible explanation of the correlation between the residuals. I think that the answer is No, assuming some locality and absence of unlikely timing coincidences. But it is very clear that with three detectors, one has a safer picture of both polarized signals, may start to do some consistency checks, and the uncertainty decreases.