The problem with these kinds of statistical snippets that papers like the FT produce, is that they are fishing for some kind of shocking/surprising easy to understand finding.
That doesn't mean that such statistics and charts actually show anything useful in a scientific sense.
The problem is that the information is presented as a scientific "truth", and many people don't have the understanding to know that all statistics and analysis involve human input (e.g. which tests to use, how to visualise, which data to exclude etc etc), and that this input introduces biases, whether accidentally or intentionally.
That doesn't mean that such statistics and charts actually show anything useful in a scientific sense.
The problem is that the information is presented as a scientific "truth", and many people don't have the understanding to know that all statistics and analysis involve human input (e.g. which tests to use, how to visualise, which data to exclude etc etc), and that this input introduces biases, whether accidentally or intentionally.