This  article investigates the correlation between population density and voting behaviour for the 2016 presidential election in the US. Not very surprisingly, Clinton won in many denser (urban) voting districts, while Trump won in many more sparsely populated (rural) districts. More interestingly, the article shows these correlations in several beautifully rendered visualizations full of details.
One example uses a lot of under-the-hood data crunching to deliver rapid information on accessibility based on travel times, selling this information via API to businesses [1,2].
The other one is a bit more static and less entrepreneurial: How far can you get away from downtown by car on a Friday afternoon, i.e. where could you spend your weekend? 
An intriguing case of data journalism, with all source code and data freely available. Buzzfeed News  has tried to identify spy planes from flight paths, with astonishing results.