Not all scientific hypotheses pan out. In fact, most don't. This is disappointing to researchers eager to learn some new truth about Nature. But it is what makes science reliable and valuable. As attractive as it might seem to have the freedom to spin out a web of scientific theory to match one's imagination, it is even more worthwhile to build theoretical edifices that can be relied upon. This is actually much better for speculative theorists. It means that there is a trustworthy foundation one can build upon, without running too much risk of wasting one's time, or even a whole career, on developing theories that become worthless when the foundations that others laid turn out to be incapable of supporting further construction.
The way that science avoids such disasters is by requiring new ideas and hypotheses to meet as many strict tests against experimental data as possible. A recent article by Frank Close, as summarized in this piece On the Nonexistence of Pentaquarks
, provides a good case study of this process in action.
Most of us would be better off in our personal lives as well if we'd learn to test our bright new ideas and cherished beliefs against known facts and data -- even if it takes some effort to acquire the relevant facts and data. And especially even though it means we sometimes have to give up on those ideas and beliefs when they don't pass a conscientious reality check.