The difference is that there were plenty of competitors whose commodity products were overtaking even the premium, business-friendly features in BlackBerry gear, just as generic PCs eventually made dedicated high-end workstations obsolete. Right now, the only credible competitor to Windows for a user-facing business PC operating system is OS X, and even then a lot more of the serious business software is better on Windows today so Apple's penetration seems to be mainly laptops for management and sales roles.
There are plenty of ways Microsoft could develop Windows that would play to their existing customer base and their established strengths; I've argued at length in previous posts that Microsoft are the best positioned business in the world to push a "private cloud" trend, for one possible example, and that they could offer genuine benefits to businesses that would justify upgrading their installations. All they have to do is regain the focus they had a few years ago and play the cards they've been dealt to best effect, and as far as desktop Windows goes, I believe that means not rocking the boat trying to be too clever and "consumery", but instead building out the technical strengths under the hood and making them easy and familiar to access.
But if their next desktop OS is another turkey, I suspect their luck will start to run out. It will take them several years to even try a different strategy, and by then I expect at least one major competing platform will have reached enough maturity to take them head on and start stealing away the big business software market. Apple could do it -- they have the technical smarts, the war chest, and the scale to pull it off -- but I don't think they will under their current management team. I suspect a more direct challenge might come from a start-up with a big name backer building something on Linux. While "year of Linux on the desktop" is still a running joke, the likes of Android phones/tablets, Chromebooks, and probably soon Steam Machines show it can work as a non-geek's OS given the right front-end. Looking 3/4/5 years from now, around the time Microsoft's next-but-one generation user OS arrived, a huge amount could have happened in an industry this fast-paced.