> So the question is, where should the fix be applied? I'd posit that there is no good answer to this question, and that attempting to find or optimize one is

> So the question is, where should the fix be applied? I'd posit that there is no good answer to this question, and that attempting to find or optimize one is a distraction. In very broad strokes, if you try to fix a problem too slowly, bad things keep happening; if you try to fix it too fast, you're liable to upset other things around you (e.g. the economy, delicate sensibilities, your short-term profitability, etc.). And the problem is too grossly high-dimensional to hope for any kind of consensus (or often, any analysis) at any level of detail beyond simply "yup, that's a problem". With no hope of finding an ideal solution, the pragmatic response is to just address the problems you can see with the resources you can afford. Trust in being directionally correct.