Scientific research cost, and usually governments fund. A government fund a scientific research for results, and for the potential benefits. It seems like a governmental investment for future. Thus, the policy suggested by the author is sound at first glance, for a wise investor won’t let money cost in vein. However, to begin with my discussion, I’d like to state two examples. The first one is Project Manhattan. Before the first nuclear test in New Mexico and the following Hiroshima atomic bombing, this whole project just looked like a black hole in most people’s eyes, consuming large amount of resource and fund for nothing. And the second is Project Apollo. When J.F. Kennedy decided to support such a gigantic project, some opponents asserted that its prospective could be “unpredictable”. That is, US government were not so clear about consequences of these projects. And we know them by now: both projects bring about some greatest breakthrough in history of mankind, and maybe more. What if the government refused funding the projects for “consequence unclear”? So here is my stance: government cannot make decisions simply by such a standard, as the author purposed. Or we may and will miss some great scientific and technological discoveries.
Unlike the laymen and rank-and-file have thought, scientific research always means risk-taking. In most cases, researchers cannot make sure what exactly will happen, though the can predict the possible outcomes. That means, the result of a research are always not promised. Some of such projects are, of course, black holes for money and resource. But it is not always the case. A scientific research is worth funding if researchers manage to convincingly describe the theories and to draw the blueprint in detail, and if the possible outcome is very rewarding. Project Manhattan made US the first country capable to produce nuclear weapon, and Hiroshima and Nagasaki bombing helped to avoid loss of millions of lives. Project Apollo made US the winner of space race, and contributed in the final US victory of cold war. Such outcomes, if partly predicted and assured, will dwarf the risk the government takes. Thus, I think a government should take necessary risk to fund some projects like these.    
Some opponent may argue that government should spend the money on some projects whose result are more promising, in order to help people in a more tangible way. They purposed, for example, “Why should NASA spend so much government funding on Apollo Project when so many poor are starving? Why don’t just buy them food with the money?” Well, in fact, the development achieved in Apollo Project did, help more poor for food. The development of space probing allowed the launch of more satellites, with which we could make better weather forecast. With the forecast, the yield of agriculture increased. Combined with the development of new material, we are able to offer more and cheaper food for public. That’s the second reason I disagree with the author. As we already know, it is often the case that we make a discovery in pursuit of another goal. Or we can simply consider such development as some by-product. However, all these following development could be non-existing had government refused to fund Apollo Project first. This is how we do scientific research and how we benefit from the research. 
All in all, I suggest a government should appropriately fund some research, even their consequences are unclear.
后记：分析一个例子就好。两个例子无法深入，也无法写完。