I've long been deeply skeptical about the notion that phones would displace PCs (I'm including Macs here) as the principal platform for computing and Internet access. In particular, I always try to bust the myth that "the next billion Internet users will first see the net on a handset" - a catchy assertion, but one backed up by zero evidence or clear rationale. I've been similarly unconvinced by Nokia's rhetoric about its smartphones being "mobile computers".
Yes, an increasing number of quick computing/Internet tasks are done on smartphones - checking email or Facebook, using a navigation app, jotting down your expenses, writing short blog posts and so on. But almost all of these are transactional updates of processes *anchored* by traditional PCs. Nobody sets up their Facebook account and configures all the settings on a phone.
I also disagree that the gulf between laptops and smartphones is narrowing. Netbooks are small laptops - the fact that a handful are sold through carrier channels with embedded connectivity is not changing the overall paradigm. Phones are "service" devices which are mostly useless without a linked subscription of some sort, while PCs are computing "products" owned and controlled by the user or IT department.
The interesting thing is that only Apple (and Microsoft to a lesser degree) seem to *really* understand the full importance and longevity of the PC - and the extra options it enables in terms of the business model for handsets/smartphones. Generations of mobile users have struggled with awful PC-phone connectivity software from almost all vendors, which have viewed the desktop as an afterthought.
Many people seem to make the mistake of assuming that Apple is targetting the 1-billion unit annual mobile handset market with the iPhone (and 4-5bn mobile users), usually citing its market share as tiny and insignificant, with a disproportionate amount of media and developer attention being focused on it. Apple itself seems happy to shrug off such criticisms.
I think this because it defines its market differently. I suspect that the real "addressable market" it's after is the installed base of PC owners/users - probably about 1.3bn/2.0bn repectively. By no great coincidence, pretty much all PC users also have a mobile phone.
Unlike purveyors of "mobile computers", "smartbooks" and the like, Apple doesn't want to substitute the iPhone against a desktop or a laptop. It appears disinterested in the iPhone being a new user's *only* source of Internet access. It wants the phone to be an adjunct - and also hopefully encourage a long-term switch to a Mac desktop or laptop.
Accessories often have some of the highest profit margins. And if they can drive the sale of premium-priced "core" products then so much the better.
But the real kicker here is iTunes. Physically connecting an iPhone to a PC means it can be totally and reliably updated and upgraded - not just in terms of the apps and OS, but also right down to the radio firmware and low-level internals. This is a lot less clunky than the over-the-air approach to device management elsewhere. The reach it gives into the installed base is hugely more than most other operator/device approaches.
I think there's more here to this. I think the notion that smartphones are somehow cheaper than PCs and will appeal to first-time users in developing countries is wrong. A second-hand $50 Pentium 3 can get a whole family online, show movies and be used for a broad set of applications. It will also last for years. (I know, that's exactly what a Sri Lankan cab driver told me in January when I asked him if he used the Internet on his phone).
In that scenario, the "cost of ownership" for a basic PC is less than $5 per person per year [excluding the connectivity, obviously]. Even a $50 secondhand unlocked Symbian device would be more expensive on that basis, given its expected working life and difficulty in sharing. There is no way that a $100-150 device would "get a whole family online" and provide anything like the functionality or useability of a PC.
Bottom line: The addressable market for smartphones is (to 80/20 accuracy) the base of people who already have a PC. The most interesting and relevant market share statistic would be the break-down of PC users who are also *active* smartphone users, with a data plan.