Agreed. It's an interesting trend. I also just skimmed over it; however, I wonder how significant is their initial TF binding occupancy data. There are a bunch of papers recently that looked at binding sites and found that they don't really correlate well with transcription. The cell is just a stochastic bag of molecules. TFs bind whereever they can. I wonder if they are just seeing noise.
Fig 1E seems promising, as it shows that the bound bases tend to be more conserved. That said, you are right that as far I could tell there were no direct measurements of transcriptional effects. Either way, it was a good read, and brought up a bunch of interesting questions to figure out next.
