Measurement artifacts or malfunction (and mistaken eyewitness accounts from multiple credible observers) are another possibility, but they're difficult to squar

Measurement artifacts or malfunction (and mistaken eyewitness accounts from multiple credible observers) are another possibility, but they're difficult to square with the data. There's just too much signal, measured in too many different ways, to handwave it away like that.