Science is real life in a controlled enviroment with repeatable outcomes. Like real life though mankind will never get to the point that his knowledge exceeds his ignorance. Science can never answer important questions and in real life we'll always have to make the best guesses with the available information. Never will there be a realistic formula for determining the outcome of marrying Laura or its comparison to Nicole. Real life activities are rarely calculable either. You might be able to determine the odds of it raining Tuesday but the only way to know is to wait.

It takes all kinds to make a real world and if we all had the same beliefs it really wouldn't function properly. Who would buy gold if everyone's a seller? The laws of nature and chaos assure such eventualities are impossible anyway. There have to be people who dig ditches and those who do arbitrage. Nature abhors a vacuum and human nature creates opportunities rather than vacuums.

There is no proper answer to the question but as a rule of thumb there are two main ways in which scientific advancement comes. The first is when accidents happen to good observers especially when they are well trained and knowledgeable enough to spot it. The second is intuitive leaps. These don't require especially good observational skills or training necessarily since we're all pretty much totally ignorant anyway.

Reason requires us to act on and believe scientific principles but, I believe, most people are well advised to use these as a framework for intuition rather than trying to live their lives based on them. We can become automatons being fed ideas as a lubricant rather than facts. Of course some disciplines require pure logic and it would be difficult for those trained otherwise to function in them. Some laboratory work as well requires pure logic and reason and can give rise to human advancement.

In the long run the most important thing for a scientist has always been his ability to observe. To me this means not only seeing what exists and seeing it properly but also examining those observations which aren't explicable. A good scientist not only is a good observer, experiment designer, and usually intuitive but also seeks data. A good scientist won't accept ignorance for an answer, merely as a foregone conclusion.

____________
Man fears the pyramid, time fears man.