The question is when to pivot and try one more time to build a product fueled by user-generated content. The author asks, Should I throw away a year of work? post 
The project has been based on a chatbot idea I like a lot. I would assume that designing "usefully probing question flows" is a new form of literature.
You've built tools for authoring and visualization of these flows. And you've designed and sought help designing multiple flows of various kinds. One writes in this literature by composing for all cases in a decision tree. The questions depend on their Socratic nature to lead the reader (answerer) to deep reflection.
I am reminded of The Pattern Language with claimed to generate solutions to problems (of design) by leading one through a directed graph of solutions. Alexander was convincing that this was possible. Your quest was to move through the graph with yes/no questions. Writing for this could be as hard, but probably not harder, than writing pattern languages.
# Experiments
I will try writing some socratic dialogs on subjects related to this question. I will imagine a process similar to one I've used to coax pattern language out of experience. See Tips For Writing Pattern Languages. Here I will only be expecting yes-no or small choice answers.