I'm not familiar with this particular set of findings, but generally speaking internal research at Facebook are generally either large scale in-app surveys (which suffer from response bias) or small scale (often ~10 participants from a single city) in-person interviews. It's useful for informing product roadmaps, but I don't think we can really draw bigger conclusions from it. Put it a different way, if this research…
Yes, if a rigorous study found that social media helped girls, and there were replications of said study, then people would buy that. This is how research works!