How do I structure genuinely ambiguous problems?

You can structure most problems now. The ones that still defeat you are the genuinely ambiguous ones, and that is exactly as it should be, because ambiguity is where frameworks run out. When there is no clean shape to find, the very skill that served you starts to feel useless. But ambiguity is not a problem to be solved so much as a fog to be moved through. The best people in your field do not magically see structure where there is none. They make a rough, honest guess at a structure, test it against reality, and then revise it as they learn. They are comfortable being wrong on the way to being useful. The framework is a tool for thinking, not a truth waiting to be uncovered. So when a problem refuses to take shape, stop waiting for the perfect structure and impose a rough one on purpose. Pick any reasonable way to break it into parts, knowing it is imperfect, and use it to start asking questions. The structure will improve as you go. Motion through the fog beats standing still, waiting for it to clear.