If not than what is the bar to show demand?
Selling a commercial software product that allows people to run MS Office on Linux isn't talk. But beyond that, if "talk" isn't good enough for you, then how exactly is one supposed to judge demand without actually producing a product to see who will buy it? And let's not forget OpenOffice, which also suggests some demand for an Office suite on Linux, a role that an MS Office port could easily fill. The obvious reasaon why Microsoft isn't interested in the Linux space is that Linux is competition, not for MS Office but for Windows. That's very similar to why Apple won't port the Mac OS to standard Intel machines. It's not that they can't. It's just that Apple serves two masters (hardware and OS). So does Microsoft (applications and OS). And the one can't risk hurting the other.