The question tossed at Shuttleworth was how important he believes Wine to be, and the role of Windows compatibility projects in general, as well as how they compared to native Linux ports that didn't require any sort of abstraction layer. His answer was short, but still very interesting. He cited the importance of Wine, but tempered that with a reminder that free software needs to “thrive on its own rules.” He claims there is a fundamental difference between proprietary software and free software, and therefore the platforms should succeed or fail on their own. Perhaps most interesting, he said that if Linux is only used as a medium to run Windows apps, there is no possible way for Linux to succeed.
While he was referring specifically to the future of Ubuntu, his words can easily be translated to any use of Linux on the desktop. As others have pointed out, attempting to emulate Windows won't do much good to capture market share. What would help is to focus on what people want or need to do, at its core. There may be difficulty in luring people away from apps they are loyal too, but software changes and is replaced as time goes on. The future of Linux on desktop doesn't lie within Windows app compatibility, but rather making the Linux desktop something people want to use due to its own merit.