The difference between liking/enjoying and gaining benefit I think is crucial here. It's akin to the difference between pleasure and happiness. No doubt the vast majority of people (if not everyone) on social media gains some pleasure from it. But as far as increasing happiness, aiding in quality of life and well-being and similar measures, it - in general - does not move the needle and not infrequently pulls in the opposite direction (in the absence of personal bonds).

The other important difference is between using social media sites as "literature" (i.e. not interacting with it but just consuming its thought-provoking and useful content, as you mention) and as a conversational medium. The former - in the absence of obviously harmful content like people showing off and then comparing oneself to them, or outrage porn, ect - is as far as I'm aware harmless and has a similar profile to reading in general.

But the constant one-off interactions, like this? It hits different. We know that consistent interaction with the same person - even if it is light meaningless interaction - has benefit. One study that comes to mind is of the daily interaction between customer and cashier every morning at the coffee shop. Believe it or not that has a statistically significant (though by no means large) positive effect on a person's mental state. Deep personal one-off conversation in person also have a significant benefit, at least in the short term.

But this - whatever this is - generally doesn't. Of course everyone is different, and obviously not everyone uses it exactly the same way, so there are outliers. But also human beings are talented at lying to themselves about the effects of things on them, often in the direction of exaggerating or inventing a benefit (and also failing at the afore-mentioned differentiation between pleasure and happiness). That's one of the big problems that researchers in outcomes studies on any number of things have to deal with. Objective observer and close other ratings of a person's happiness after say therapy or anti-depressants are often wildly less rosy than the person's self-report. We humans are often not good at being honest with ourselves.

But you are right about the "attention economy" context. We - near worldwide it seems - have little perfected the art of leisure. Our institutions have failed us, and we in turn have created little to take the place of the old ways. Our choices are too often between similarly poor options. That's probably why work and over-work have remained as popular as it has - what lies outside of it is too often stupefying. Which is one reason why I'm here now, talking with you, embracing (or stomaching) this momentary pleasure in what is perhaps a trade-off with a bit of my happiness.

Google scholar is the easiest source, but I can help by sharing a few (there is no single paper I know of that sums everything up). It's a difficult topic because so few solid studies have been done - but everything we know suggests that its effects are - on average - weak. In other words, online time appears to neither add to nor detract from the lives of most people (at least directly). It's just time filled up with vanilla.

See: https://www.sciencedirect.com/science/article/pii/S2352250X2...

And: https://www.annualreviews.org/doi/abs/10.1146/annurev-psych-...

https://europepmc.org/article/ppr/ppr407158

But there are a minority of folks on either end of the spectrum. On the negative side, specific types of personalities, attachment styles, and so on can suffer significantly on the internet and especially on social media. The media loves these studies, there’s a ton and they are easily found so I’ll just refer you to scholar.

Then there are groups of people that gain from the internet. The few studies that have bothered to investigate how people use their time online and in social media in detail have found that talking to internet randos in one-off fashion does not help, but that - and I’ll quote here:

“When people use social media in a truly social manner (i.e., actively interacting with meaningful social relations in a way akin to in-person social interactions) it was positively associated with psychological well-being. We propose this is because truly social usage promotes meaningful social relations, which result in positive psychological consequences such as reinforcing one’s identity, feeling valued, and mitigating stressful situations. Yet, when people use social media in other ways (e.g., passively engaging with weakly connected others, celebrities, brands, companies, or strangers typically for entertainment purposes) it does not influence psychological well-being. Therefore, how and how much people use social media has implications for their psychological health.” (in https://www.msi.org/wp-content/uploads/2021/05/MSI_Report_21...)

Of course this is but one study, and an as-yet unpublished one (though it appears well-designed). A meta-analysis from 2017 using this as a frame though backs this up here: https://journals.sagepub.com/doi/full/10.1177/09637214177308... - and a more recent though poorly written one generally mirrors it: https://www.frontiersin.org/articles/10.3389/fpsyg.2021.6787....

Another recent study - this time on the 60+ crowd - found “answering questions online were positively related to depressive symptoms” and looking at photos of non-family members on social media was associated with anxiety (https://journals.sagepub.com/doi/pdf/10.1177/146144482110543...)

Then there’s the work of Alexander J. A. M. van Deursen, in which he focuses on who exactly gains from their time on the internet and how - and how this in turn perpetuates inequality. You can find his work here: https://research.utwente.nl/en/persons/alexander-jam-van-deu...

But the foundation of the argument for me lies more in the research on what contributes to happiness and well-being, and which of those variables can be realistically gained from time on the internet and social media. What in the long term will make a difference in a person’s life? And invariably the most impactful outcome is the formation of an offline friendships and so-called "social capital". Which appears to generally be a rare thing - https://www.tandfonline.com/doi/abs/10.1080/1369118080263543... - though it does happen, and it seems self-disclosure is a key to that (among other things) - https://journals.sagepub.com/doi/abs/10.1177/146144481985874....

There’s more research that can be brought to bear on this, but that at least may provide a start for your own inquiry.