I know of a multimillion dollar pilot plant built on a gas chromatograph peak. Oops!, it was discovered when the plant was almost done that the peak was something other than what it was thought.
Can you say egg on your face? Some managerial (and others too) people wanted something so bad they took a short cut. Well, it was only money.
The best engineering in the world can't fix bad science. I worked on the testing of the tile repair systems for the shuttle. We engineered a backpack device that could place ablative material in a damage site on the orbiter in 0-g, vacuum, operated by a crewmember in the suit. We could get the crewmember to the worksite with the backpack, perform the repair, with contingency mechanism for all the "just in case" event we could imagine. The decision was made higher up, over my objections, to parallel path the ablator material science and the engineering and operations development. In the end, the material bubbled in vacuum, and peeled out of the damage site under reentry dynamic pressure. The project was finally scrapped because the material didn't work. Beautiful engineering, ruined by bad (or non existent) science. And don't even bring up the issue of money....