they did ask an interesting question in the text that detailed this award:
"According to research organizations such as Evans Data, KDE is the most popular desktop environment. How does that square with the fact that GNOME is the default desktop of one of the most popular distributions (Ubuntu)? We have no idea." - Linux Journal
here's my educated guess (or, amazingly long rant):
first off, ubuntu is not actually one of the most popular distributions (sorry distrowatch observers). in my travels i certainly don't find it to be dominant outside of the vocal online community. it's an extremely promising player in the global linux market and is growing at a terrific pace (i'm starting to see it pop up more often in deployments) but it still lacks the ubiquity of red hat, suse, debian, etc. i both use and like kubuntu, so i have no ax to grind here.
let's sidestep the ubuntu popularity question for a second, though, and address a more generic question: "how much do the software choices made by operating system vendors (OSVs) matter in the open source market?"
i believe kde has several assets when it comes to influencing usage decisions that offset OSV defaults.
the power of community, particularly when it extends across diverse demographics and interests, is immense
perhaps our largest asset comes in the form of human resources: we have a terrific community of developers and users that form a self-organizing global mesh of voices that fairly consistently and quite accessibly support and promote the use of kde technologies in locally and globally relevant ways.
these individuals choose for themselves who they talk to and what "market segments" to address (even if they don't realize they are making such choices). this results in people who are appropriate for a corporate audience talking to companies while those who are NGO friendly talk to social change communities while those who are more university oriented ..... well, you get the picture. this happens without them getting in each other's way; in fact, it often serves to reinforce everyone's message and efforts.
kde delivers technologies that 'speak' strongly to most people; thousands of people working on hundreds of software projects that are personally relevant results in software that is tuned for people with similar needs and desires
some of kde's assets are undeniably technical. there are many applications and technologies in kde that simply rock the house. some speak to the media generation, like amarok and digikam. others speak to the sys admins of the world like kiosk. others speak to artists such as krita which has been getting a nice amount of attention with the release of koffice 1.6, and deservedly so as it's a very exciting project. some speak to educators like kalzium and kstars. and on and on.
people may try something because they heard good things, but they tend to stay when they discover good things. better yet, they tell others about those good things. kde has good things.
culture is intertwined in all things used by humans. kde invests in culture and society because we appreciate the value inherent in them and this in turn makes kde attractive to human beings
some of kde's assets are cultural. we tend to be accommodating of others, both technically and socially, while more than willing to share what we have. we have an immense population of translators who help make kde culturally accessible to billions of people on this planet who prefer something other than a handful of languages that originated in europe. we also have people that show up on a regular basis to events around the world to meet with people, talk with them, listen to them, share their personal interests and passions, invite them in to our cultural circles and who join their cultural circles.
now, i'm not saying kde is perfect or that other free software options have nothing to offer the world. au contraire! if that was true kde contributors could go on permanent hiatus and soak up the rays on a beach somewhere instead of working every day to improve things. we have technical, social and structural improvements to attend to that keep us very busy. and i'm happy that people have options besides kde so that those who like other approaches can still use free software and be satisfied without us having to mutate kde for them.
there are also some prevalent but questionable assumptions out there. for instance, users do try software out when given the chance. just because they get something by default when they install a particular operating system doesn't prevent them from trying out other things.
case in point: t.'s mom is looking into desktop linux right now with some of her online friends. they are surveying the landscape and trying a lot of things out. some are choosing gnome while others choose kde (and i'm sure some will make other selections such as xfce). they are deciding for themselves what they prefer. why? because they can. the only cost is some bandwidth, a blank cd and little bit of their time. for that small price they get to choose for themselves and that freedom is valuable to them.
there's also the implicit assumption that the distributions that really "matter" are the global ones. having travelled a bit in the last couple of years i've seen enough regional distributions and free software support efforts to have decided that this simply isn't true. local and global often stand on equal terms when there is true freedom, and many regional efforts (both of the OSV and 'grass roots' sort) make kde an important part of their strategy.
one of the most revolutionary things about free software is how it allows for profitable companies to exist, project planning to occur, grassroots social movements, personal choice and individual involvement all at the same time. it contains for many of the best parts of a "top down" society without sacrificing bottom-up control and participation. it thrives on diversity rather than optimizing towards uniformity and grows in direct relation to openness and human vibrancy all while creating new opportunities for science and business alike, two endeavours which have become increasingly linked with the diminishment of openness and humanity over the last century.
to me it is the fundamental strength of the free software model which explains why kde succeeds, regardless of people's expectations or predictions. put more simply: theory be damned when there's evidence at hand.