Slightly visible fenestra margins and other externally discernable skull landmarks do not neccesarily make a piece of dinosaur palaeoart "shrinkwrapped". As long as one doesn't make the animal look emaciated and give the impression of unrealistic "sinking" of the soft tissues into fenestrations by overdefining the outlines, there are actually more options to explore than one may realise. Due to the change in the soft tissues directly covering the skull between the fenestrae/fossae and other surfaces, it's not totally unreasonable to speculate that bony margins margins would express some kind of subtle surface landmark in some cases. In other cases, it would be equally reasonable to illustrate a total lack of such landmarks.
Ultimately, there's no single convenient analogue for how we should be restoring the facial subcutaneous tissues of dinosaurs in areas unassociated with robust integumentary correlates. Take a look at Cathartes aura (the turkey vulture), where the skull morphology is quite discernable in life. Then, compare it to other closely related cathardids with rather similar skulls, where one will observe excessive tissues largely obscuring any bony landmarks.
I believe that there may be a bit of a greater misconception throughout the palaeoart community regarding what "shrinkwrapping" really is, and when it's actually a real problem worth criticising. Maybe I'll talk about this in more detail more at a later date. In short, what I would define as "shinkwrapping" has much more to do with the commonly illustrated absence of various soft tissues that can reliably inferred to have been present (often at a deeper level than integument and subcutaneous tissue layers), and far less to do with the conservative reconstruction of more superficial/subcutaneous tissue volumes.