From my initial glance at the post and paper, they seem to be claiming that the factors they do look at account for almost all the variation. If this is a corr

From my initial glance at the post and paper, they seem to be claiming that the factors they do look at account for almost all the variation. If this is a correct attribution, then there can't be big role for any other factors, known or unknown. Now, it's easy to fool oneself in this sort of modelling exercise - by fitting lots of parameters to the data, one can seem to account for lots of variation, without actually capturing the true relationship. I haven't looked at the paper in detail yet, so I can't say whether or not they avoided this pitfall.