There has always been discussion about how the Linux world is confusing. One aspect has always been distributions - there are so many choices, when someone says Linux or wants to switch to it, how should they know what form to switch to? However, this to me seems like fundamentally the wrong question. People don't care that Windows uses NT (which NT do I get? Vista, 7? XP?) or Mac using FreeBSD. In both these cases you have the same fundamental issue - a fairly common base (NT, FreeBSD) with some modifications and changes for a particular use case. Slap some branding on it (Windows 7, Mac OS X) and that's what you promote.
Of course, with a culture of developers and technically orientated people, the fact that a distribution uses Linux may become important, the same way as NT or FreeBSD being used might do, and it's not surprising that conversation falls back to discussing Linux as a single product like Mac or Windows, but this is a conversation and PR mistake. When that's not the primary audience you're selling too, you effectively put the cart before the horse. You're setting yourself up for confusion, complication and marketing and discussion failure. Ultimately, people rarely (if ever) use Linux on its own, it's part of a larger collection of various projects which then make the whole. Linux is important, but where in the chain you discuss and bring that point up - if ever - is and should be different depending on who you're talking to.
I think this is valuable to think about from a developer standpoint too. Do developers who create Windows applications or Mac applications talk about how they're developing for NT or Unix/FreeBSD? Certainly many Linux distributions are more similar than they are different, but to continue thinking of them as the conglomerate of Linux that if you target one you must target all I think is naive, and is partially due to this culture of talking about it in this way of Linux as the operating system, rather than through the prism of the distribution.
At the moment this may seem most sensible due to market share statistics, commonalities and worries, but I think as we start seeing certain distributions seemingly pull away from others in audience figures, we'll start seeing the conversation change from developing for Linux as a whole, to developing for Android (heavily modified Linux), or indeed even Ubuntu. Linux will be a common base kernel that these operating systems use as their starting point, but it'll be more akin to the way Mac uses FreeBSD - not in the sense of things like contributions back to the original project, but more in the sense of marketing and as targets for developing applications. It will still carry advantages for standards across these as open standards always have, but it'll be less of a worry for developers who only want to target the largest current potential audience.
Though this all becomes less of an issue as web based applications become more popular. This is also less of an issue for free software development in general where targeting one particular distribution doesn't necessarily lock out the possibility of others, in fact making it likely given enough interest in the project. However, this is something that regularly comes up when proprietary land tries to cross over, which I think has more to do with their model of development than they sometimes think.