Apart from the 55% human accuracy, which apparently comes from a completely unrelated study, the bit that really stands out to me is it reports the accuracy onl

Apart from the 55% human accuracy, which apparently comes from a completely unrelated study, the bit that really stands out to me is it reports the accuracy only drops from 72% to 68% when controlling for demographic accuracy in the US (a little more noticeable in the UK). Considering demographics alone gets you 60-90% accuracy on voting intention for many US demographics, it strikes me as extremely odd it would have have so little impact on the model.