Could it be that the light was recorded moving that fast, but that it is somehow an illusion based on how the data was measured?
On the other hand, they're measuring a particle that nobody else has measured before, so it could also not be measurement error. That's the interesting possibility.
I want to see where this goes.
You can also filter by time - you create a source and expect a surge of detections shortly after. In this specific case, they saw the surge before they expected it, indicating the neutrinos traveled faster than they should.
Look any direction in the night sky. Do you see a star. The odds of not seeing a star in the direction you're looking are far greater than the odds that you don't have neutrinos coming from that direction.
You can't shield against neutrinos, you can't currently detect a neutrinos position at two points in spacetime. Haven't yet read the details about this if they repeated several times with the same result then it looks interesting - have you a link to their paper (to save me looking!) thanks.
My first thought was that I wonder if this might be a simple case of super-luminal group velocities or something similar.
I didn't look at the data, but the point in time the detection cuts off is also relevant - it should match the time it took the first neutrinos to arrive at the detector. If some of the original neutrinos interacted with something outside normal spacetime and that made new neutrinos appear closer to the detector, some of the original neutrinos should have arrived at the detector at the predicted time.
Anyway, this is a very interesting concept. Hope you can somehow test it.