People often believe that by simply mastering existing theories and concepts that somehow they will assuredly discover or create something new. Though mastering previous theories and concepts is necessary, it is not sufficient for developing something new. It is one thing to master that which already has been done, it is quite another to discovery/create something new.
The problem is that it is not possible to rigorously derive new theories from old theories, in other words it is not possible to discover/create something new thinking strictly in terms of old structures of thought. For example it is not possible to rigorously derive Quantum Mechanics or General Relativity from Classical Mechanics, though it is possible to derive Classical Mechanics from QM and GR. So, the process of discovery/creation can't be too rigorous simply because the new theory contains ideas and concepts not contained in the old; you can't extract concepts that are not implicit in the fundamental assumptions of the old theory. Sometimes these new ideas may appear irrational, if reasoned from the perspective of the old ideas. A strange inversion occurs the old ideas become irrational when the new ideas are shown to be valid. On the other hand, the reasoning can't be too loose, it must be possible to extract the old theory from the new. This is why high intelligence does not guarantee discoveries/progress; the path to the new theory is not simply a complex derivation from the old, in other words it is not linear. To be fair innate ability does help, though lack of innate ability doesn't imply a lack of discoveries/progress.
Obviously this implies risk because the ideas have not been shown to be valid, the chances you will turn out to be wrong are high. The reason why we take this risk is because the old theories sometimes prove to be inadequate, this is not because previous physicists were incompetent rather, it is because we have information/results/insights they didn't have. Sometimes the new information/results/insights shows there is a problem with their theory. There may be elements of truth in the old theory, but these elements are constrained by the information available to their creators.
Another problem is that people may not want to question old ideas/theories on the basis that old ideas/theories have worked so well and are so well reasoned that they must be true. Sometimes people don't question old ideas/theories on the basis that if people "smarter" than themselves couldn't improve upon the theory, neither could they. At other times people believe that everything that could ever be thought of has already been thought, so what's the use? Sometimes people believe something to be impossible hence they don't consider it further. Some people believe it is not their position in life to develop something new, that is left to "special" people. Others simply don't want to make the effort. Finally any new idea brings the risk of the unknown and of being shown wrong hence, many people do not want to take the risk. The entire paradigm of thought weighs heavy on people's shoulders any attempt to cast it off is difficult.
Many people wish for a life or career of guarantees; guaranteed success, guaranteed position, guaranteed wealth, and for the physicists, guaranteed truth (or approximation thereof). The problem is that there is none; you may spend your entire life working on a theory only to have it turn out to be wrong, you may be the most intelligent person on earth working on what appears to be the correct direction for a new theory only to have it destroyed in an instant.
These prospects sometimes lead people and sometimes entire cultures to rarely take risks, to rarely think of something new, or to even pursue new avenues of thought. According to them, it is much safer to work with theories and ideas that are already established, especially if one desires guaranteed results or success. Sometimes entire societies inadvertently stifle free thought, especially if the cultural structures of thought implicitly impair the development of new ideas. This usually takes form as a belief in certain unquestioned assumptions that may or may not be evident, sometimes it is a belief that every idea must be rational (in common parlance it must make sense), the only problem is that what is considered rational is constrained by available information/insights/results.
To sum up it is necessary to master established theories but it is not sufficient to make new discoveries.