My reasoning here is that although a selection of personas may map to the profiles of key website visitor stereotypes, when you start to look at them in more detail you find that your users are all different. So whilst at the very highest level your personas are very different, at the lower level (e.g. the completion of a certain task) a subset of these have alternative needs. And to confuse matters, these alternative needs are shared with users in other personas.
Let's just consider 4 different personas of a typical online service and overlay a few different user profiles over them. As you can see, the personas can cover the main functionality a service provides, but each user has a place in one or more of them. And in the case of some really important users (such as those who need accessibility compliance) they could be present in all 4 personas.
Or put another way... there is no point creating specific personas for each user type. You'd end up doubling your persona numbers just accounting for accessible users if that was the case.
No, that doesn't make sense.... and in this way the developing of personas really doesn't map to something you actually need on your digital project. An understanding of what functionality you need to prioritise over other functionality.