My interpretation of that (especially over a short period like this) is that sleep restriction causes swelling in the grey matter, which caffeine reverses. Interpreted like that, caffeine would be protecting the brain, not harming it.

I think it's far too early to assume assume the "swelling" is a net harm or that preventing it is good in the long run.

Perhaps it serves an important protective or regulatory purpose which we haven't yet identified, because we still know so little about sleep and consciousness.

An analogy: There are lots of systems in your body that try to prevent and deter you from casually walking on a damaged leg, and we can override some of them with various external drugs, but that doesn't necessarily mean it's a good idea.

So it very well could be that thinking on a worn neuron, trying to do memory access on it, might damage it even more (leg analogy)? It could be possible that the "swelling" could be redirecting brain to use other paths but with caffeine you're utilizing a worn-out neuron.

Maybe with chronic deprivation + caffeine habit, you're weakening your cognitive function - repeatedly using a misfiring neuron. At least, with no caffeine, you'd feel your memory is hazy and hard to recall, go to sleep.

IANANeuroscientist, but my idle-thought was that a lack of sleep could cause a buildup of waste byproducts, and structural adaptations might either reduce the effect of the buildup or else make its removal marginally easier.

Yes, that's my point. You can read this either way. But we do know that sleep deprivation is harmful, and it's conceivable that this might point to a mechanism by which that harm is caused. Or it might be a defence mechanism which the caffeine is now interfering with, making things worse.

What about the imaging method makes you interpret it this way?

How did you make that connection? I didn't get that info from this study, but I have seen other studies that suggest sleep deprivation causes neuroinflammation, and I have seen studies that say coffee can reduce neuroinflammation. I have seen that the neuroinflammation described as a defense mechanism to assist with the removal of toxins. Possible connection being that if coffee reduces neuroinflammation when there are toxins present that may be a cause for more damage? I have zero qualifications, just trying to understand.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9658704/

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5476750/