Over the past several years, Microsoft has made several attempts to lower the cost of using Office, typically paired with buying a certain computer. In the past, limited-functionality versions of Word and Excel were paired with advertisements (mostly to purchase the full versions of Word and Excel) and the installer for Office 2010. Other arrangements have included Office Online and now, copies of Office 2013 Home and Student or subscriptions to Office 365 Personal are bundled on various types of devices.
What has long seemed odd to me is that Office is included or bundled in some manner on the (physically) smallest devices, and nothing but a trial is included on bigger devices, even when both devices have the same amount of computing power, display the same number of pixels, and are very similar in price. Somewhat counter-intuitively, I would argue that Office Online makes most sense on the smallest of "desktop" and "laptop" devices that have the least storage. However, without either introducing new features to the suite, upselling OneDrive storage on its own, or adding a way to stream the desktop applications back in, there's no way for Microsoft to monetize this right now. In addition, the smallest devices are also where Office Online makes the least sense, because they're the ones likely to be without connectivity, as their users take them to class, meetings, cafes, and generally "around."
The Office situation on the smallest computers is sort of weird, because they're (almost) all licensed for the desktop Office suite, but the least expensive systems (like the HP Stream 7 and the WinBook TW700) don't make sense to use Office on, if they even have enough disk space on which to install the apps (and many of them just don't). The fact that they're bad computers to run Office on doesn't prevent it from being there, of course. For disk space reasons, these are the machines on which it makes most sense to use web applications, or partake of application streaming and virtualization on server platforms. However, these systems are the most likely to suffer a total loss of connectivity for an extended period of time. Even the Hewlett Packard Streams 13 and 8, which have T-Mobile WWAN functionality, are not a guarantee that you'll be able to get to your data at all times.
With the departure of Office Starter 2010, users of "normal sized" computers (anything bigger than 13-inches seems to be the demarcation) are down a choice of how to use Office. I don't like that Office comes free with computers that are too small to productively use Office, and you get nothing on computers with larger displays and better keyboards.
Using Office online could still make a lot of sense on an inexpensive stationary (or even "mostly stationary" as many 15-inch laptops really are) computer, such as the Bay-Trail based Inspiron 20, or one of the bigger laptops, if only for cost reasons. These systems also still have slightly faster processors and more memory to make running the web aps easier. The challenge is that such a machine also definitely has what it takes to either run regular Office365, an old perpetual copy of Office, or even OpenOffice or LibreOffice, and there's not much in Windows itself to encourage the use of Office Online.
I think what may be happening is that Microsoft would far and away rather just sell you a subscription to Office365. In fact, I would be completely unsurprised to find out that the Inspiron 20 has the bits to Office365 already loaded, waiting for you to sign up and sign in. The unfortunate (but clever, nevertheless) likelihood is that they're going to do almost as little as possible to inform you about Office Online.
One other thing I wonder is whether or not Microsoft could be working up to another reduced-functionality desktop Office product, and what the cost structure for any upcoming Universal version of Office for Windows 10 will be. When it launched on iOS and Android, Microsoft required an account to use the Office apps, but have since opened up some of the functionality to OneDrive-only users. (It's worth noting that you can still buy OneDrive quota on its own, free of an Office license, even if it's well established that Offic365, with its unlimited storage, is a better deal if you're in it for the storage.) Without knowing what the cost of the Office "Universal" application for Windows 10 will be, the situation today is still that if you buy a really inexpensive 15-inch laptop, you'll ultimately pay full retail, which can cost more than the computer itself did for a perpetual license.
Granted, relatively few home users have a whole lot to gain from buying the perpetual edition of Office, especially when it does not include additional OneDrive storage. Monthly or yearly pricing on Office365, which includes all of the core Office applications (including Access and Publisher) should be more palatable for home users, who will start to look at it as a utility, or simply get a new license each year courtesy of a cheap, small computer, which doesn't necessarily have to be the one running Office.
If Microsoft wants to compete with Chromebooks, something's gotta give. A large part of the reason the inexpensive Windows PC category exists is because there is said to be demand for cheap hardware on which to run legacy applications. There's no compelling reason to recommend a Windows system to somebody who is just going to install Chrome and use Google Docs because Office is too expensive and nobody is aware of Office Online. Users who do know about Office Online can use it just as easily in Chrome on a Chromebook as they can in Chrome or IE on a Windows computer. The slim nature of ChromeOS might even mean that on hardware in the same price range as any given cheap Windows system, Office Online will run better. The connectivity issues remain, however, and most ChromeBooks are direct competitors (physically, computationally, and price-wise) with the machines that include either Office 2013 Home and Student or Office 365 Personal.
Unfortunately, one of the biggest solutions to this problems, which is just to make network connectivity better so that users of "Stream" class PCs can use Office Online and not waste their local disk capacity, isn't something Microsoft and the PC OEMs are currently positioned to do. The other possibility is that the Universal version of Office is going to be a lot lighter on disk space and buyers of "big" inexpensive computers will accept the operational limitations of these applications, whatever those may be.
One of the other options I do want to see for Office is a return of some way to "stream" the desktop applications. I think App-V may be the best way for Microsoft to do this. It'll let users map their local storage for use with Office, use their OneDrive storage, and RDP (which is how App-v applications are delivered) is a much quicker and more efficient way to get to these apps than the old streaming solution, which involved installing an IE plug-in and downloading the whole application, and only worked on x86 PCs running Windows where you had administrative access to the computer, so that you could install the IE plugin.
If I had to make a guess at how cheap or free Office licensing will change over the next few years, it's that the smallest computers will come with the Universal version of Office for free. It'll be licensed perpetually, which will be an advantage over the current Office365 Personal subscriptions on many tablets and netbooks. The pack-in incentive from Microsoft's part on these systems can move from being Office365 to a cheaper-for-Microsoft OneDrive storage quota increase such as what you get with a Surface 2 or HP Stream desktop, and Microsoft can offer Office365 "proper" licenses on bigger computers. Though, even that may not be strictly necessary depending on how Microsoft licenses those Universal apps. If those are free for all copies of Windows 10, (the way that you can install Word & Co on any iPad for free) then you could argue that this takes care of the "need" to provide an inexpensive way to get Office.
A potential advantage to this is that it will encourage that type of customer to save their data online, but it also lets you save data locally to your computer. I have yet to spend enough time with it to find out if it allows users to save information directly to SharePoint sites, or open documents off of SharePoint. Microsoft allows subscribers to different products different access to features within the iPad and Android applications, which is a convenient way to upsell people on things that not everybody uses regularly, such as change tracking. Microsoft could hypothetically use in-app purchases to allow users of the apps who do not want an Office365 subscription to purchase additional or new functionality (such as, say, the citation manager).
I believe Microsoft's strategy so far is to rely on the Office mobile applications to sell Office 365 subscriptions, and the marquee product in Office 365 is definitely the desktop Office application. Each Office 365 subscription level defines a number of computers on which you can run the product concurrently, but the thing Microsoft has neglected to do, as far as I can tell, is put any actual limit or indicator in showing how many mobile devices you're using it on. Either it's officially unofficially unlimited, or Microsoft still considers the mobile applications to be accessory products, or access conduits, rather than productivity applications on their own.
This isn't necessarily bad, but it's definitely different from Apple's productivity strategy: When the iPad was new, each of Apple's productivity apps were $10, and the whole suite on the Mac cost $50. Now, the appropriate version of these applications is included with any given Apple device.
I think that as with the Office Web App, Microsoft is getting its hands into as many platforms as they can, but the question is how and when Microsoft will directly monetize these programs. The New Microsoft seems really open to different arrangements and possibilities, but not so much so that they're going to let an entire platform of wallets get away for free, but I also doubt that even if they officially allow installation on more than one mobile device, mobile- or web-only users are going to be willing to pay $6.99 monthly or $69.99 yearly for document storage and Word on the iPad.
The next possible option is that Microsoft is going to add more functionality to it. The space is fairly flexible, and with the OneDrive application, other applications can send data to the space, and you can open files from OneDrive in other applications on mobile devices. I suppose the question is whether or not Office365 is the best productivity solution on mobile devices, and if it is, what it's worth to users.
It'll be interesting to see how Microsoft moves forward with licensing a variety of things. Office as a service to consumers and small businesses is still relatively new, having been released three and a half years ago at the end of June, 2011. Over the course of those years, Microsoft has added some new products and features to the suite, and increased the allocations of features, such as storage space. We have yet to see if Microsoft will move to this model permanently, and when that will happen. For the time being, they seem to be testing the waters. Maybe they were looking when Adobe (much to the continued, active dismay of their customers) did away with version numbers and announced the end of perpetually licensed copies of their core products.