I do appreciate you linking to articles but these articles are pretty terrible.
re #1: The WHO adding internet addiction reflects the amount of sway China's political processes have over WHO declarations more than anything else. Their internal political narrative is that this is a problem and the WHO is being used to support that. The repetition of the falsified blue light hypothesis re: sleep is also informative re: the quality of citation #1. As for "changes in glutamatergic and gabaergic" signalling... if that doesn't happen when it means you're brain dead. Glutamate or GABA expressing neurons literally make up ~3/4 of the neuronal cells in the brain. And both are regulated extraceullary by glial cells too. You cannot do anything without changing this. If they'd done fMRI or PET or something and could shown long term abberant changes in the glutamergic signalling in the shell of the nucelus accumbens then maybe it'd be saying something. But they don't and I'm getting ahead of myself.
Citation #2 shows that when people are doing something relaxing and then they stop doing it they aren't as relaxed. That's hardly surprising. The arguments seem to be pop-sci level characterizations of the brain where any change is seen as significant or having a valance, good or bad.
And then they go and cite obviously false out-dated concepts like the idea of dopaminergic cells being neccessary or sufficient for expressions of pleasure/reward,
> Dopamine plays a critical role in this circuitry, for the subjective pleasure associated with positive rewards, and the motivation or drive-related reinforcements associated with eating, drinking, or drugs [73,74]
>The initially pleasant, so-called rewarding effects of the drug are relayed by the release of dopamine in the nucleus accumbens (NA) by the synaptic endings from the neurons of the ventral tegmental area (VTA) of the mesocorticolimbic circuitry [79,80].
It's actually glutamergic cells in the shell of the nucleus accumbens that are necessary and sufficient (but not all encompassing) for pleasure expression in mammals. Dopaminergic neurons can be blocked off with antagonists and the expression is still complete. The modern understanding is that mesolimbic dopaminergic populations encode for wanting and reward prediction. Glutamergic cells encode for reward/pleasure. I'd hope that someone writing a policy paper like this would cite up to date knowledge but it is excusable and a side point.
The real problem with #2 is that it doesn't actually talk about withdrawl symptoms in "digital addicts". It talks about widthdrawl symptoms and neurochemistry known in actual drug addicts and then just implicitly applies that all these statements must apply to the behavior "digital addiction" too. They don't show data about "digital addiction" withdrawl.
The third article is behind a cloudflare wall and I cannot access it.