Perhaps, but relying on best current knowledge that is backed by evidence is a pretty good policy. It may not always be correct, but it's a whole lot better than blind conjecture. Even if it's a guess made by the most educated guy in the room, that's still the most educated guess we have. Hindsight is 20/20, and we bias our view of past decisions with current knowledge.

It's good in certain situations, when there is a good body of scientific knowledge to fall back on.

The tragic flaw of science-based policymaking is that scientific understanding requires evidence, and evidence is only available after an intervention is put into place. Because of this, if you rely on science alone to dictate policy, it becomes a reactionary force that opposes all new ideas as too poorly understood to be implemented.