If you’ve got a written description of the thought process by which you came to the idea, keep that! But the thing that should be published should be the thing that is that plus supporting evidence like citations and logical reasoning describing how such a thing could have come to be the case. Simply don’t destroy the evidence, and what you’ve got is pure improvement. If the non-rational hunch and analogy part seems hard to fit in to the cited polished product, then keep them as separate docs with links to each other.
If you’ve got a written description of the thought process by which you came to the idea, keep that! But the thing that should be published should be the thing that is that plus supporting evidence like citations and logical reasoning describing how such a thing could have come to be the case. Simply don’t destroy the evidence, and what you’ve got is pure improvement. If the non-rational hunch and analogy part seems hard to fit in to the cited polished product, then keep them as separate docs with links to each other.