Obviously it was not known from the beginning that it would be a dead end. However, engineers have this blind spot for keeping track of bigger-picture objectives, as opposed to technical specifications. If you set out to build a lawn mower, and end up with a rubber duck, this will be deemed a failure. But if you do build the lawn mower, all is good. Even if it costs a million dollars, and all the lawns in the country remain unmowed (except for a handful of government properties). So long as someone is paying for continued development, where's the problem?

In other words, a lot of such government-backed projects utterly fail in their objectives, not so much due to lack of prophetic foresight, but due to inability to re-evaluate when it becomes clear that the previously chosen approach can no longer lead to the envisioned outcome. ITER is another fine example of this.