The authors are data scientists. So articulating what exactly the pitfalls are that disqualify them from working with an epidemiological dataset would be really helpful. As a non data scientist, it is not obvious to me what qualification they are lacking when it comes to working with an existing dataset.