This post discusses how to streamline the other way - extracting code from a fragmented and longish SE answer but I imagine a two-way workflow like:
Fashion an answer/question in an SE-templated notebook with a submit button that converts both text/input cells into their SE equivalents. The notebook could also be effectively uploaded to SE by having the conversion include a link back to a public repository to which the original notebook is also sent.
NotebookPut@ ImportString[ Uncompress@ FromCharacterCode@ TakeWhile[ Flatten@ImageData[Import["http://i.stack.imgur.com/mV4FJ.png"], "Byte"], # != 0 &], "NB"]
$\endgroup$