Well, I have a headache now. I had forgotten how poorly written some of these published papers tend to be. Anyway, sorta long summary after skimming a few of the studies and that meta-analysis:

1. The meta-analysis worked through data obtained in a range from 1967 to 1995. It found that the consumption of "explicit pornographic material" appears to create a mostly consistent change in the behavior of adolescents and measured in four categories. I'm on mobile so I won't go back and grab those categories, though the participants are mostly balanced between them. It should be noted that this analysis is trying to push a hard need for practical findings in our modern day despite only taking information from the range provided. Moreover, there is an air of bias regarding the findings.

2. Other studies have concluded, in general, that while we believe there is an increased risk of early sexual development and even deviance, it has been difficult to replicate these consistently.

3. Most studies conclude that modern consumption of media by teenagers may or may not increase the risk of deviancy many of us would consider stereotypical risks that teenagers take.

Basically, science is struggling a bit to show a positive correlation. They think there might be something there, though looking at research into other types of media you'll find similar findings.