I agree but for a slightly different reason: I think the notion is also highly subjectively applicable... some types of problems really do need a lot of foretho

I agree but for a slightly different reason: I think the notion is also highly subjectively applicable... some types of problems really do need a lot of forethought to make efficient - waiting until later can cause significant problems when decisions are hard to pivot away from or undo. An easy example is DB design, the cost neglecting these considerations upfront can be massive and sometimes impossible to migrate away from.