The linked article is all about prediction of reward. Not reward itself. The dopaminergic populations of neurons, in the VTA or not, are not necessary for expressions of reward itself which is mostly in glutamergic cells in the shell of the nucleus accumbens(s). But the dopaminergic VTA neurons do project to the accumbens as part of reward prediction. I say this so we can avoid reviving outdated 1970s dopamine as 'reward chemical' ideas.

>“It seems like dopamine is signaling not just for [reward] prediction errors, but errors in general.”

This makes sense. More evidence that dopamine isn't especially related to reward.

> which is mostly in glutamergic cells in the shell of the nucleus accumbens(s)

i'm not really convinced this is any better fit for "reward" than other explanations. What evidence is there that such a reward system exists in the first place?