for me personally one thing i much prefer about observable over Jupyter as an example is that the cells are automagically understood to be part of a "DAG" (https://en.wikipedia.org/wiki/Directed_acyclic_graph). In other words... instead of my jupyter workflow of rerunning cell 10 then 15 cuz it depends on the output of 10, then back to 9 so i can update this... then trying to remember if i should go rerun cell 12..... i don't have to do any of that. it's all magically handled. that's my favorite "unique" value prop.