Double-clicking is much more difficult to do correctly, and distinguishing repeated single clicks from double clicks is another minefield (manifesting itself in

Double-clicking is much more difficult to do correctly, and distinguishing repeated single clicks from double clicks is another minefield (manifesting itself in e.g. the mouse gesture needed to rename a file on Windows - you must single click it once to select, and then single click the name again to edit it, but if you do it too fast, it's treated as a double click). The real problem IMO is that we have effectively standardized on two-button mice as the baseline, with all UX then designed around that even though many mice have 5+ buttons these days. The three basic actions that desktop UI has ultimately converged on are: select; activate; show additional actions (context menu). These should logically map to three independent buttons, such that the two most common actions are mapped to buttons that have fingers resting on them in neutral position - e.g. left = activate, right = select, middle = context menu.