Such good questions! I’ve recently been thinking that UX begins with user expectation – how someone thinks something ought to work – and then moves toward whatever we can do to make things work to meet those expectations.
UX studies can let us know about many expectations, but not all are knowable and even if they are sometimes knowable, there may be reasons why, gasp, there are excellent reasons not to meet the them. (At least in every instance. Such as no scrolling. Ever.)
I find myself thinking often about the friction between the way things work in the moment, the way we’d like them to work, what can be learned by the pain points and how many pain points could we and should we remove?