It is common knowledge (or perhaps conventional wisdom?) that the spread of information technology into all phases of society has occured, at least in part, because of the pervasive success of many Microsoft programs, mostly remixes of well-established prior art in the areas of operating system and "office-type" applications. They have been less effective using the same patterns of "innovation/piracy" in the era of near-universal networking.
The "killer app" has taken shape because of continued fairly radical innovations in the area of making the largest presences due not to "enterprise" activity (which we will see becoming obsolete through bloat/inefficiency/boredom), but to individualistic undertakings. The means for anybody on the planet to become something like Amazon.com was shown clearly when an individual hacked a new form of interaction so that his friend could more easily engage in the hobby of collecting Pez dispensers: eBay.com!
What started as "Web Pages" to show off one's photos/ideas/preoccupations morphed into the most talked-about/participated-in area of the Web: the Blogosphere.
When wikis began evolving it became clear that collaborative editing was about to change forever and the Wikipedia established the supremacy of descriptive vs. prescriptive definition - as well as the possibility of internationalization/inclusion due to accessibility of the medium AND the message.
The talent pool (their users, not their techno-teams!) who will furnish role models and nexi for most of us to learn the reflexes of using the "killer app" will be central to the success of the proposed re-distribution of the emphases on prodution. The creation of a really huge user network will make a hierarchical support group totally ineffective/inefficient because of the Web vs. tree reality of it all. The profit center will be in bell/whistle/update/upgrade sales and possibly orderly subscribed-to maintenance of aspects of the network.
Although almost all of the current application packages are central to communicating, the integration of them via "the answer" has been sadly neglected because, in part, there is a tendency to regard it all as part of the operating system when it is actually a function of the creative process - in other words, a Content Management System. Essentially ALL content prepared for and obtained from the network has been created using numerous already-widespread tools, perhaps principally word/image processors. The techniques for "managing" this content, particularly during its creation, has been more bolt-on than integrated-into methods. This is clearest in the case of accessibility/usability which are almost universally treated as after-market modifications rather than as integral aspects of the creation process. A truly effective authoring system will more or less force such things as device-independence and indexing to be simultaneous with authoring.
The "how" of it has always been blathered about within Microsoft where there has been a so-far-ineffective effort to make all aspects of these processes share much commonality, but the notion of a CMS being in charge simply has not found traction. For various technical/historical reasons this can be done most effectively by a slight restructuring of priorities within the organization.
The "why" of it is largely because the productivity gains that will come from having individuals with the power previously monopolized by institutions will make problem-solving of the sorts typified by "how do we feed the starving babies?" and "how do we make competition collaborative rather than combative?" much more prone to solutions before this old geezer shuffles off.
As Bucky Fuller makes very clear we are all billionaires due to the inexhaustible resources of universe (gravity/light) and the singular Korzybskian metaphysics achieved through time-binding. Exclusivity preserved through weaponry will be of little interest to golfers/weavers/singers when each of us has more real wealth/power than the kings of yore.
Love.