I think the practice of heavy question editing, especially after answer(s) have been provided, especially if the edit contains new facts that the questioning person (or user(s) that have already answered) might not have known, should be discouraged.
The user who is tempted to do such heavy edit with new facts, considerably different structure representation or partial answer should rather provide a comment or answer.
UPDATE, Semi-hypothetical example:
An experienced chem.SE user sees following structure image
and is temped to replace it with nice:
(which won't even be considered a heavy edit).
However, there's already a comment
You are ignoring strain in a cyclobutene ring …
and the question author might have been unable to realize that. So the redrawn structure with fundamentally better quality and color (which are irrelevant), should preserve the original fine double bond graphical representation detail:
The information content of the question should not be altered by the question edits.
(Inspired by What is the stability order for these octahydrobiphenylene structures?)
Of course, I am not talking about questions converted to community questions.
My opinion will be probably wrong if chem.SE (*.SE, SO) should be considered rather e.g. knowledge base, than reputation-oriented question/answer service.