I'm not a statistics expert but. . .I think a potential issue that's being overlooked is that the statistical levels of confidence that are being used for a lot

I'm not a statistics expert but. . .I think a potential issue that's being overlooked is that the statistical levels of confidence that are being used for a lot of these big data analyses are not that tight - they are what social scientists use as opposed to what physical scientists and engineers would use. So the correlations are more likely to be spurious. Or not. . .