I think improved longevity over time is relative to non-caffeinated solar cells, caffeinated solar cells last longer (they don't improve over time)
It's specific to perovskite solar cells, though, which aren't in commercial use because their lifespans are so short. They're one of the 'holy grail' high-efficiency technologies like magnesium batteries, so a 13x increase in lifespan is at once hugely important and not enough to make them realistic.
My assumption was that since the slope looks fairly flat, this was still a nonstarter for most applications - 86% after 2 months is good progress, but commercial standards are 80% after 20 years. Even so, it's a big step up from the material's past of failing outright after short spans.