The real issue has never been about the best platform, it has been about the most effective business tactics. Time and again, Linux products have been launched, only to be outsmarted by other players with a more forceful (aka less ethical) approach to doing business. How many of us remember Corel Linux or Lindows? Do you really think the Asus netbook revolution died because the Linux versions sucked?
Why do you suppose the leading proprietary Office suite vendor made occasional forays onto competing platforms but always half-hearted and never longer-term than the heat of battle demanded? Or why did they chase Western pirates for licenses but not (until very recently) the Chinese? Surely it couldn't have been an astute move to get around China's cavalier attitude to licensing by pursuing volume to chase off Linux before eventually monetising the market? Oh, come to think of it, that's exactly what they said it was at the time.
No, what Linux needs is a whole new ballgame, a world in which we don't want fat clients and stultifyingly familiar desktops any more because we have moved on. One where it can innovate and compete on a level playing field. How about maybe mobile data, digital convergence, cloud services, HTML5, .... Heck, it doesn't even matter if we call it something else - Android maybe, or ChromeOS or WebOS or Firefox OS or whatever. Some folks will no doubt be reinventing the old GNU/Linux environment alongside those reinventing the even older Windows (TM) environment - possibly with about as much success, we shall see (Unity anybody?).
Currently around 80% of computing devices sold into the new mobile world run one flavour of Linux or another - mostly Android. ISTR that represents around 60% of the total (fixed plus mobile) market. The Year of Linux was 2012.
This year we get to rub salt in the wound.
[Update] And I think Valve may have a view on how best to do that.