Unfortunately you present a rather cheap and obvious strawman here. The word significant is used very strictly to describe the outcome of a statistical test, in the context of a mathematical framework and dataset. You know effects can be significant and very weak too, like with detecting gravitational waves or detecting wind speeds by looking at the minute deflections of a telephone pole in the breeze. Anyone with statistical experience in research knows that it is not so easy to find statistical significance where there is none. Rejecting the null hypothesis is only easy if you make the wrong statistical tests. If everything is correlated to everything else, try checking the statistical significance of the number of lakes in a country to its population's life expectancy. Good luck with that!