"It's possible that the same thing is happening in this group of research as well."

In this case we already know that the most important mediating variables for predicting outcomes on the relevant measures are parent-child talk and self-directed play, so it's very easy to see how each hour of watching TV causes increasing damage. I'm sure some of the gap is because of mere correlation, but it's also clear that TV watching itself is a huge part of the problem if only because it prevents the child from getting the types of interactions they need for their brains to develop properly. (whether or not the TV itself fries their dopamine system or whatever directly is still an open question.)