Any empirical evidence to show any of these things are actually good? Some "seem" to be common sense (e.g. less form fields to lower sign up friction) but a lot of the other things appear to be what one UX designer thinks it's a good idea rather than anything that might actually help you.