The difficulty with learning WPF is not so much the API as the model. It's a very different mental model than you'd use with something like Windows Forms.
Rather than writing methods that imperatively populate a UI element, you generally data bind to properties on an object. To get complex behavior, you generally use some level of composition.
As an example, if you have a series of items you want in a list, with a piece of text followed by an image. In Windows Forms, you'd get the list and iterate it. For each item in the list, you'd create the control for the item, and the text, and add the picture, and then add the new subcontrol to the list. (Exact procedure may vary based on the type of control. You may add SubItems instead of just making one control, etc.).
WPF would handle this very differently. In WPF, at the top level, you'd declare a container object and bind it to the list.
You would then give that container a template to use to display its items. The template is basically another control, which defines the positioning of sub-elements, which are bound off of an instance of the class that the list is populated with.
It ends up with a very compositional feel, and is absurdly powerful. It's also a very different model than most developers are used to, and until you can internalize the model it's very common to run into problems trying to do things the way that you would in Windows Forms/etc.
Once you get used to the model, though, going back to other APIs is painful. Not because they're suddenly hard, but because you know how easy things can be.
