You are completely ignoring the MS Office market. Try typing an essay or a report on an iPad or making a good presentation. You certainly can, but its not even remotely efficient (or a good "user experience" or "magical").
These days the only time I need Office is when some idiot sends me a document in one of their proprietary formats. But taking your point to be the more general "word processing / spreadsheet " market (since I don't feel like the data base aspects really caught on mainstream) I observe that there are fewer and fewer documents that are 'word processed'.
What is perhaps more relevant, is that I have tools available on the iPad that are at least as functional as using a typewriter to type up something (the pre-Office world) And if I absolutely must type rapidly on the iPad there are keyboard solutions out there. But I don't think I am compelled by the argument "You can't do Office[1] therefore you can't get rid of your PC."
What I do see is people using a device for all of the things they used to use a full blown computer for, knowing that those same people never exploited half of what their computer could do. And much of that has been driven by the notion that there is "one thing" which you use to do this stuff. A wedge was driven in that notion by 'smart' phones, and a bigger wedge is being driven into that notion by tablets.
[1] "Office" here is of course a stand-in for a variety of capabilities which one might classify as 'information manufacturing'
Project Managers and Business Analysts are another category of people who rely on MS Office heavily as well.
Programmers may not.
I don't think anyone expects that general purpose personal computers are ever going away, but there is a pretty compelling argument that the bulk of computational assets will fall into the 'appliance' category rather than the 'general purpose' category.
As Atwood points out, Microsoft strove to have a PC in every house, and they succeeded! PCs have higher penetration in homes than TVs. However, given tablets and their capabilities, it is looking more and more likely that this trend will stop and that the number of homes without a PC in them will begin to rise again. its an open question how many homes that will be but it seems likely it will happen.
There is little stopping them from using newer tools but inertia.
They don't need Office. They are just used to it, much like accountants are used to their 12C's
Now most students are being trained to use Office at school.
I use to see the world from an "Open Source" programmer's point of view: "You have the choice to use OpenOffice". But then after I saw the reality for a while, things start to sink in and it becomes obvious that people don't want to be liberated the way I thought they ought to be.
People are OK with MS Office. It becomes a piece of software that, no matter how hard geeks want to argue about the Pareto Principle or not, will stood the test because people were being trained regardless its complexity.
So yeah, from "feature/functionality" perspective, there are alternative. But from skill-wise, willing-to-learn aspect, there aren't until Universities and/or schools start teaching young people to use something else.
OTOH another argument is: you don't need the other alternatives if you have Office.
A person close to me is currently taking Accounting courses and calculator + paper do not cut it. You need Excel.
EDIT: I don't care for MSFT, I loathe Word. But Excel on Windows is a genuinely fine tool -- overused by a lot of people b/c the Windows anti-pattern keeps them from proper tools, but still a really useful piece of work. You can't do detailed modern financial analysis without it. The alternatives -- Linux, Google Docs, even Excel on Mac -- aren't even close.
What some courses teach is not financial math - it's "how to get numbers out of a 12C". This is just perverse.
It's limitations are outweighed by how fast you can do cut and dice data, including pivot tables (ie, basic OLAP). Google Docs is still too limited (ie, try filtering and sorting on multiple columns).
If you know of a way I can do this in a desktop or web-app for the entry price of Excel (free would be nice, but I don't mind paying), I'd be happy to know.
Tablets are good for certain use-cases, but when it comes to typing to save your life (i.e. your business), they are simply not a match and will likely never be. Couple that with the small screen real estate (spreadsheets these days are enormous), and you see how the iPad can be taken seriously only by the detached upper echelons who play golf all day. The rank-and-file will keep asking for (lighter) laptops forever more.
Take a step back, and really consider how likely it is that laptops are FOREVER. It might be true, but things change and I think it's historically much more likely that ANYTHING you point a finger at today will be transformatively replaced in due time, than it is that you'll be able to accurately predict a permanent future for whatever an ultrabook is.
I actually don't have any idea what an ultrabook is, but it sounds like a very small idea... not something I'd bet on or invest in for the very long term.
Touchscreens are fundamentally limited by the size of your finger. Firstly because you have to physically press the UI element you want to interact with. Secondly because your finger obscures whatever's underneath.
Of course, you can get around this by decoupling the physical location of the UI element and the physical location of your finger. It's called a trackpad.
Touchscreens are great for some tasks. But they are not great for tasks involving high information densities (writing, spreadsheets, mostly anything that can be classified as "editing"). For those, people will continue to use keyboards and mice until they evolve cone-shaped transparent fingers.
As soon as they start to get over the 11-to-13'' form factor (and lower in price), they're going to replace the current generation of work laptops. They'll get thinner and thinner; as batteries and bluetooth-like technologies improve, sooner or later they'll even lose USB ports, and you'll be left with a device that is literally a screen and a keyboard. Some models will have a detachable tablet screen, no doubt, but the heavy lifting will still be done with a real keyboard.
I don't think it will change tomorrow. But I could see the home market making the shift, gradually. Where the business market will be fifteen years from now we'll have to see.
1. Every office in the world will buy each employee a separate office room so they can dictate their documents without being disturbed by the other employees.
2. Every office in the world will buy each employee a $10 physical keyboard.
I don't think most folks would end up working from home 5 days a week -- there are still things that work better face-to-face. But I could see folks _only_ coming in for face-to-face time.
Now it becomes a hassle to carry all of these docks and accessories around with you so you think "gee , I wish somebody could invent something that combined my screen, keyboard and mouse into one unit..."
Another point to bear in mind is that whilst people will have computers at work and at school a great many people , especially students will want to do work from home.
People do not just base their requirements on what they need 99% of the time, but rather on what they forsee their possible needs might be.
Let's take cars for example, how many people do you know that own medium to large size cars with 5 seats and a trunk where for 90% of their daily driving the car only contains the driver , possibly one passenger and an empty trunk?
To exploit your analogy (cars are always good for that :-) how many people own trucks these days? There is always a market for a truck, its the best selling unit of Ford's inventory, but people who only occasionally move stuff or carry stuff often will borrow or rent a truck when needed rather than own one)
The other interesting part of this conversation is the conflation of cause and solution. Your point 'especially students will want to work from home' is particularly salient. When I was a student the PC didn't exist, I still worked fine from home :-) But more importantly as the presence of PC's in the home grew, teachers gave students the opportunity to do work on them rather than say writing it out long form, etc. But as tablets become common at home and with students, and especially tablet based text books, the opportunity to provide even better tools for students is possible.
Imagine a world where your history home work is to write an essay on some topic in the text book that is also on your tablet. Writing an essay is pretty trivial mechanically, you don't mouse a lot and you can do the whole thing even from a soft keyboard.
And this comment for me really sums it up : "People do not just base their requirements on what they need 99% of the time, but rather on what they forsee their possible needs might be."
This is spot on, and there are many people who are looking at tablets and deciding that it can cover even their foreseeable needs. And that, for me at least defines, the term 'Post PC', when something other than a PC can meet the extant and foreseeable needs of someone with respect to information communication and manipulation.
I think the precision and speed advantage you get from a proper keyboard and mouse setup vs a touchscreen, especially when you consider how cheap both of these items are will mean that they are around for some time even if it as an optional add-on rather than a core part of the whole interface.
I see the "post pc" as more of a rethink of the PC minus the general historical baggage that various versions of Windows etc have carried with them for last 20 or so years.
What I think people are really clambering for is a better user experience from their software.
Let's take the filesystem as an example.
Years ago it made sense to have an OS with the concept of files and folders being exposed to the user.
Why? Because you needed to be able to group your stuff together, either things for a particular project or because they were of a particular file format (word documents , spreadsheets etc).
You would also have "drive letters" which told you whether it was on the "C drive" (inside the computer) or the "A drive" (portable). This is no longer important since we can trivially sync most things to the internet regardless of where the bits are physically stored. In fact Unix doesn't even have the concept of drive letters at all.
File formats can now easily be replaced by Mime types and meta info since fast SSDs and indexes can be used to search either the device itself or cloud storage very quickly. This means it is easy to have an app that loads and immediately provides a list of all the documents which it can work with.
The only requirement that is left is the "grouping", folders are pretty bad for this because you can only group in one dimension. A better system might be some form of "tagging" for example tagging as "work" or "my project" etc.
By removing the filesystem from the user's consciousness you have already massively improved usability and I think this is probably the biggest thing that has made the iPad so easy to use because it essentially turns the install process for an application into a 1 stage thing. Combine this with having updates automatically install silently in the background and an application ecosystem built around the idea of the internet and web as a first class citizen in the OS rather than an extension of it and you have a very seamless experience.
The physical "input into the device" landscape though I see as something blown open and I really don't think there needs to be one dominant paradigm here.