As I understand it: they have found possible causes of error, and the original results are being discounted until a re-run, which is due in May.

Yeah, last I heard was that they didn't take into account the relativistic effects of the satellites moving. The satellite timing system used to clock the neutrinos was thus off by the same amount that the neutrinos were moving faster than the speed of light. It may be coincidental, but it seems to explain it. Funny that relativity proved wrong a result that was supposed to prove relativity wrong.