I originally drafted this post a little while ago and a lot of things have changed since then. The datacenter Apple are building to support their expected iCloud revamp of the MobileMe service and likely integrated iTunes functionality is of staggering proportions. There are in all likelihood only a few dozen comparable systems on the face of the earth at this point in time, though I expect they are going to become more prevalent in the next two decades. Cloud computing, going by the common moniker "The Cloud" which I still dislike as a misleading nomenclature, is a genius idea. The accessibility and potential cost benefits for most of us are simple enough.: you don't need significant storage capacity or the best bleeding-edge hardware in your home to enjoy the benefits of both; rather you delegate that responsibility to someone else. Someone who essentially owns those things on an epic scale on behalf of you and however many other customers they can snag. It's happened in business before, we're all familiar with the terms 'dumb terminal' and 'thin client'. Centralised server-bound storage and number-crunching processing power is not a new concept at all.
The difference now is the internet. What has previously been a far more situational solution to computing needs thanks to multi-megabit connections and more sophisticated software has the potential to be the future's primary computing paradigm. Even more importantly it can be modular, competitive and task-oriented, ideal grounds for a thriving, cut-throat and therefore consumer-beneficial market. You may choose iCloud for your music, Google for your email, Flickr for image processing and Microsoft for your productivity software. Each application being administered and run out of a different datacenter. Your home machines would simply perform the task of integrating those services into a coherent end-user experience.
Again this is already happening: the App Store or the Android stores are to a greater and lesser extent simply platforms for integrating disparate services into your handheld device.
It does feel like a number of companies are being very tentative with their steps into this arena and some, particularly the media industry, are outright resistant. Google has the best foothold with it's suite of software and will be making another giant leap with its Chrome OS based offerings later in the year. I see a need for compromise however. It's not always convenient to 'Cloud-source' everything. What would you do with no internet connection? A business has to continue despite service interruptions. You still want your music on a plane or on the London Underground. Sometimes the data contract on your mobile device just can't handle your demands without astronomical expense. Most cloud services will need at least rudimentary device-native offline support. It's unavoidable as we still need our autonomy and independence from the internet even if we choose not to exercise it most of the time. There really is no one-size-fits-all.
It remains to be seen how well Apple understand this. Their MobileMe offering was solid (after early teething issues were ironed out) but never really seemed to understand how it should integrate into OS X. Their iTunes offering could suffer from a similar problem, with music collection sizes running into the tens and hundreds of gigabytes it's use could be onerous and expensive. iDisk still feels disconnected from the OS X user experience when it should be the default storage location for all user created files with seamless application support and integration such that it's as smooth to operate as your machines built-in hard drives.
I'm hopeful though, Apple seem to have the best demonstrable understanding of the practical use of internet enabled personal computing. Where Google and Microsoft seem to represent the two extremes, Apple is happier somewhere in the middle of the spectrum where experience can be curated according to practicality and not ideals or outdated business models.