Status of FCAPS model? Useful? Obsolete?

Someone at Forrester research wrote an article in 2003 that said FCAPS was an obsolete model because
  it was conceived during a time when mainframes were in use. I haven't read the article but the
  premise of it seemed a bit overboard to me.

Does the FCAPS model still hold currency among network managers/engineers today? Do the functional
  categories it uses make sense of the management activities of modern networks? Is there a set of
  activities that in and of itself warrants further categorization? For example, should project management
  be added to the model since most well run networks emanate from adherence to generally accepted
  project management processes?

We can all think of examples where a piece of network technology does not map neatly into the OSI model.
But because enough of what goes on in networking does map neatly to that model, it is still useful to refer
  to it. I believe the same is true with FCAPS. I see too that companies like Cisco still refer to it in some
of their Networkers presentations.

As I understand it, the purpose of the model is to outline the overarching categories of activities necessary for
  the successful management of a sizeable network. The ultimate impact of failing to manage these areas
properly is loss of productivity, revenue or opportunity. Outside the realm of profitability, the FCAPS model
is a teaching device whose purpose, like all models, is to explain phenomena which otherwise would be
a confusing, jumbled mess. Does the study of FCAPS profit a person if they intend to manage well a network?
Is there some better model worthy of study? If so, why?

Sean had some comments on the shortcomings of FCAPS for carrier networks awhile back. Any fresh comments
  are welcome.