I have a question concerning the general trends about GUI development in small-to-big software projects having a polished GUI as a primary requirement, both from the esthetic and usability POV, over the Microsoft platforms ( both traditional desktop and "metro-like" applications ).

More specifically, many new GUI technologies ( XAML as the foremost representative ) sell themselves as promoting separation of GUI authoring from GUI "logic".

Now, my question is: to what extent does this actually happen in the real world ? that is, how often software teams using these technologies actually employ people solely dedicated to the authoring of the GUI ( = people that reads or writes no code in any language other than those supported by the authoring technology itself, like scripts or the markup itself ) ? when it does happen, do the project leaders make a trade-off between production costs ( possibly minimized by the said GUI-authoring/GUI-logic separation of concerns ) and flexibility ( what if, say, you need a graphics effect or behavior not supported by the GUI technology ) ?

thanks for answers !