Why does having a strong linear correlation not imply causality?
Correlation does not imply causation - it just means two variables are linked.
Linked variables doesn't mean that they necessarily cause each other. It could just mean that they are caused by the same thing.
For instance, say there is a statistical link between immigration and crime. It is not necessarily that immigrants are criminals, but that immigrants are typically poorer than non-immigrants, and poorer people have more reason to turn to crime.
Of course, in some instances, immigrants are criminals, just as non-immigrants are criminals, but simply from the correlation we have no grounds to form a conclusion.
We also don't know which variable causes another. It could just as easily be that crime causes immigration.