>I've found that versatility plays a huge role. I did some contract work where I was brought in primarily to build some complex reports in Crystal. But my skills in T-SQL wound up bringing more value, so the scope of the relationship with the client expanded significantly.

Undoubtedly one of keys to my success has been maintaining a diverse skillset. I've expanded, or at the very least remained valuable to many an account this way. It also gives me the ability to go after many more accounts than I would otherwise have access to. However, that having been said, I believe there is also something to be said for those in IT whose target market is a very specific niche. Ultimately if you or your company are successful and have future prospects to retain that success or even grow it, I'd say stick with it. Introducing a bunch of risk factors into a business model that is already working must be justified by the potential gain. Just as the only time someone (I'm speaking from a business standpoint here) should expend the resources necessary to become adept in another area of IT is if it will mitigate enough risk to justify the expended resources. Justification is based on the goals of an individual or company and I accept the fact that not everyone is as interested in taking over the world as I am. For some businesses and developers this definitely means that they should be expending some resources in the area of .NET and becoming proficient in creating solutions with it, but for others it only introduces risk with no discernible benefit.

Put another way, while it is true that the market for .NET is far more vast, MS IDE's for .NET are improving at a much faster rate, and the future potential for .NET appears to be very bright, it doesn't mean a hill of beans to an independent/corp. working with another language/framework/paradigm that has a workable business model, sustained growth, and future prospects well into the future. Conversely, if you are an out of work developer, experiencing a sustained decline in revenue due to a drop in demand, or can reliably forecast a significant decline for your current skillset in the next 5 years (ratio of supply vs demand must be considered... a drop in demand with a similar drop in supply will cancel each other out), you could be a fool for not considering .NET as a possibility and you are a fool if your reasons are based on biases you have against it or Microsoft in general, or you're just plain scared of change.

In the end, extremes should be avoided, the truth lies somewhere in the middle, and bad advice is just that, no matter the source.