If you are concerned about reading on devices, you may find this fairly extensive summary in Scientific American of interest. I guess the article comes down slightly on the side of paper, but does admit that the findings seem to be shifting over time (more recent studies are less likely to show differences). Experience may have solved early problems with reading on screen and the devices may have become more “eye friendly”.
I find some of the explanations baffling (biological interpretations that suggest we treat words on paper differently) and some explanations accurate but mostly due to a lack of experience (the ability to move about within text easily). I am also interested in the opportunity for multiple individuals to highlight and annotate the same content either for private use or for sharing. There are just some new opportunities that technology makes available.
I do think some reading activities that would clearly demonstrate the value of device-based reading exist and were ignored. For example, the ability to easily search within multiple sources and the ability to search for personal highlights or annotations. My interests are somewhat different from a casual reader. I have not purchased a physical book in several years but dozens a year for devices.
I suppose my bias here is obvious, but the article does offer a slightly different perspective if you are looking for arguments in favor of paper.