What historical examples fit the rule that an idea is not stupid if it works?
Short answerThe examples include improvised fuse waterproofing, the Channel Dash, and naval ramming, with uncertainty over whether the waterproofing example came from SOE work or early contact mines.
NotesOne example uses prophylactic kits to waterproof time-delay components, although the recollection is uncertain whether this belonged to SOE work or early contact mines. The Channel Dash is another candidate because its success followed a plan that sounded extraordinarily risky. Ramming appears self-destructive and can leave the ram lodged in the other vessel or broken off. A later explanation recalls an acid-operated delay that eventually released a detonator. The recollection may conflate devices, but it illustrates the search for a reliable delay before microelectronics.