The Doppelgänger Effect (updated)

Regulars here, such as there are, will have noticed that the look of my blog has changed a bit. Like many, I started with a default MT stylesheet ("Stormy"). But after encountering one too many blogs that looked just like my own--call it the doppelgänger effect--I've begun to chip away at the stylesheet, for now more chromatically than structurally.

Lev Manovich makes the interesting argument that because digital objects always operate within a finite (however vast) set of parameters, that a truly radical gesture of originality is to simply retain the default settings. Strictly speaking, though, he's talking about applications-based interfaces, and not stylesheets. Indeed, with the rise of CSS, skins, etc. there is now a more pronounced division than ever between "form" and "content"--something that first-wave interface revisionists like Brenda Laurel argued strenuously against. So, I'm wondering: are there genuine aesthetic possibilities that get lost in the neo-fundamentalist separation of styles and data, or is the old organic ideal just a kind of tired Romanticism?

Update 6/10: Here's a relevant passage from Laurel's Computes as Theatre (1991):

Usually we think about interactive computing in terms of two things: an application and an interface. In the reigning view, these two things are conceptually distinct: An application provides specific functionality for specific goals, and an interface represents that functionality to people. The interface is the thing that we communicate with -- the thing we 'talk' to -- the thing that mediates between us and the inner workings of the machine. The interface is typically designed last, after the application is thoroughly conceived and perhaps even implemented; it is attached to a preexisting bundle of 'functionality' to serve as its contact surface.

Laurel's own theories of HCI, as presented in the book, are aimed against this then-prevailing view--but the interface as "contact surface," an add-on to a "pre-existing bundle of functionality" is precisely the relationship between data and styles that's been reinstated by CSS, is it not?

To some extent I'm playing devil's advocate here, since I cut my teeth on SGML, the spiritual ancsetor of the current XHTML/CSS paradigm. But I would like some help in thinking this through. Note too that the separation of data and styles goes against the grain of the old humanistic saw about the mutually informing and inextricable nature of the relationship between form and content (got to go back to A. C. Bradley for that one).