not if that analysis requires more than trivially installable programs. what good is a notebook that uses cortex, for example, if a user doesn't have cuda installed? if they have knowledge to install and setup cuda, cortex, and learn clojure, they probably have some preferred environment, being it emacs, or cursive, or something else. notebook might be of convenience, but not that important imo.

another thing that i think misses the point is attributing python's and r's success to notebooks. while those are certainly helpful for spreading the word and attracting attention, it is not the main reason people use python and r. the main reason for both is, simply, that there is a lot of software available for them. so, not the notebook, but ggplot, tensorflow, and an obscure analysis tool X that you need but would never understand how exactly works internally. that's what brings the cargo, not wooden airplanes on the earth. speaking of clojure, those are the things that are currently lacking the most. there is (almost) nothing to display in the clojure notebook.

this is almost org-mode notebook level https://www.youtube.com/watch?v=PRlFJSf3xqU cc @elise_huard