I think social media probably isn't great, but the data also isn't particularly conclusive, and much of it is very broad and unhelpful. For example, in the first study you linked, it had a sample of 143 University students accross three different social media sites, measured only using app battery usage. While it isn't to be completely discounted, esspecially alongside the larger body of work, it's sample size is tiny, and the data collected isn't esspecially representitve of what is supposed to be measured, as it includes non-social-media tools like Facebook Marketplace while excluding social media accessed via the browser (among other flaws). The vast majority of studies on the topic, regardless of findings, have similar flaws.
IMO, social media as a whole is too varied to meaningfully lump together, and a lot of the impacts that are commonly attributed to social media are better attributed to deeper causes regarding lack of sociatal trust and support. Tracking the impacts of social media is hard enough, nonetheless the impacts of something like the lack of third places, and impacts of economic inequality on mental health.