Once the need for Office goes any, then any operating system will work for many people. Linux has finally won in the consumer space in the form of Android. Hopefully, we'll see a huge flood of "desktop/notebook" Android/Chromebook devices. Get "Office" in the cloud and the 90% desktop monopoly that Microsoft has had for two decades finally goes away. Don't get me wrong... This isn't Microsoft being the wrong company to have a monopoly, it's just silly that any company has a 90% market share. We don't buy 90% of our cars from General Motors.
Let's get a little more innovation in the consumer space. Google, Apple, Microsoft, and maybe even... Ubuntu dividing up the market will be better for everyone.
Office is completely bug-ridden and unstable, and always has been.
I haven't opened up Office since 2009. And the only reason that I opened it then is that I was getting attachments from Office docs from users who used it to communicate stuff with me that could have been more effectively communicated in plain text emails.
OEM preloads (look up the Novell / DR DOS lawsuits for specifics, there's some wonderful quotes), and destroying WordPerfect and Lotus 123 through a combination of a bundled, inexpensive suite and killing both competing products in the DOS -> Windows conversion were key.
You neglected to mention file formats, the Outlook-Exchange hegemony, Active Directory, and various proprietary integration tools (Visual Basic, OLE, C#, SharePoint).
And it's probably worth noting that Microsoft did cultivate and create a vast ISV ecosystem with a huge list of titles serving a vast array of business needs. If the OS and Office ubiquity established its dominance, this cemented it.
But since 2005 the cracks have been growing visible wider. Starting then, execs at most firms I've worked for or with had Apple laptops, not Windows systems. In some cases, the CFO/bookkeeper's Quickbooks workstation was the only Windows machine in the office, in others some nontechnical staff used Windows but all of engineering and many others were on either Mac (mostly) or Linux (some). And application use is moving increasingly toward SAAS offerings, especially for engineering and project management (wikis, bugtracking, project management and planning tools, etc.), but also increasingly for office tasks (documents, spreadsheets, presentations).
The one and only version of MS Office I've ever owned was Office 97, and for the most part I've avoided having to use it for more than very incidental tasks either personally or for work ever since.
Most of my text processing is in vim. After far too long playing with it, I've finally started picking up LaTeX (it's pretty amazing, largely straightforward, and the results are fantastic). 'awk' is my go-to "spreadsheet" tool (I prefer code/data separation), though Gnumeric and LibreOffice offer very nearly anything Excel can do (I can understand how an Excel jockey might not care to make the switch, but the tools are largely as capable and interchangeable).
Isn't that pretty much the attitude that RIM had as well? This is not a successful strategy when you have to compete with a new class of devices.
PCs and mobile devices have overlapping functionality, but the use cases and contexts in which that functionality is applied are very different.
By conflating the two markets together, and allowing (poorly-conceived, IMO) product strategies for mobile devices to spill over into their flagship PC software products, MS has made a very serious error.
Microsoft isn't cool, and that's okay. Microsoft is the boring boyfriend you know is always there and will always load Office documents easily (see 'rayiner below). It's not sexy but it doesn't have to be.