NOTE: The conclusions drawn from this poll are not indicators of the industry situation, but are merely personal thoughts related to the number of voters who participated.
It is time to bring in the results of the How important is cross-platform development to you? poll. Its results were no surprise to me, to be honest. I've seen one too many situations, where developers couldn't care less whether their application can be cross-platform compatible, and there is no one to blame for that. First of all, we have to understand that in a lot of business situations, there is a well-defined user environment, and developers are aware of it. So when it comes to writing applications, it is easier to convert the "outliers" - people running a different environment, than spend time creating a product that can run on multiple platforms.
The results look like this (out of ~32 total votes when I last checked):
As you can see from the chart above, 44% percent don't consider cross-platform development that important. The software world is in a very divided state, and a lot of platforms offer their own benefits for developers who create solutions particularly designed for an environment. For example, if a developer creates an application that integrates with the Windows 7 taskbar, there is no sense to even try porting it to other platforms because those lack the necessary UI elements. Don't forget that we are also talking about OS-based performance optimizations and platform-specific toolsets. Look at DirectX and at the amount of games written specifically for Windows and you will understand what I am talking about.
33%, on the other side, consider cross-platform development being very important. These developers are most likely working with Java (stereotypical assumption, but it's out there nonetheless), Mono, Python, Ruby, core C/C++ or are focused on web development in the first place.
The other 20% of developers that put cross-platform development in the somewhat important category are the people who are willing to dedicate some extra time to experiment with existing solutions and see how those can be implemented across multiple systems. I can put myself in the same group of developers, as I am yet to get to work on a project that requires hardcore portability across various OSs - all work done by me in the cross-platform field is on a completely enthusiasm-driven basis.
The last 7% of developers that participated in this poll consider that it is not worth the effort investing in cross-platform development. There might be many reasons for this, including the fundamental inclination to work with a platform due it being the "comfort zone." Also, the reason I mentioned in the very beginning - there is simply no incentive in trying to implement a solution that can work on different systems.
Overall, it seems like cross-platform development at its core is going to be a bit stagnant in its old definition. Java and Mono are nice, but there are nicer options out there, especially with emerging HTML5 that will allow developers create applications that target not only desktop, but also mobile platforms, that do not support third-party portability frameworks.