Perhaps the most useful aspect of thought experiments is to reveal the limitations of our reasoning ability. We tend to assume abstractions, yet when we think carefully those abstractions disappear into a myriad of fine detail. For example, people babble endlessly about "consciousness" yet no one can properly define it; worse yet, however we choose to define it, studies conclusively show that we're partially conscious occasionally and largely unconscious most of the time. We can construct thought-experiments that elucidate this unfortunate reality as well as conducting lab studies to demonstrate the same thing. The beauty of the thought experiment is that it can trigger ideas for empirical studies, the results of which are the final arbiter. Unless we want to escape into solipsism like small children hiding under the dining-table.