After 10 minute read here's what I get out of it: The study is a meta study that doesn't actually test individuals but uses data from other studies that do tes

After 10 minute read here's what I get out of it: The study is a meta study that doesn't actually test individuals but uses data from other studies that do test individuals to assess average group performance under sleep deprivation (either total or chronic restriction) using an algorithm that tries to use the other studies as inputs and predict their results to draw out patterns by finding dosing strategies that optimize performance. The advice seems to be: reduce total caffeine amount because otherwise recovery to normal is worse, take more caffeine on days with less sleep, and less on days with more sleep, delay the dose to closer to sleep period ( otherwise your performance tanks as your approach sleep period, i.e., you only stay high for so long ), and if you can don't dose on the first and last days of your sleep restriction period, just go "natural".