not so much...
You seem to confuse failure to innovate and improve with actual economic failure. They are quite different.
Linux is great in the innovation department but fails badly on the economic one.
M$ is making as much money today as ever and despite the whole hype about Apple surpassing it they still have MORE NET INCOME at the end of the year, and by several billions.
Microsoft will not try a new strategy for their OS line until the money REALLY starts to go down, when they REALLY start to lose market share. Do you remember the netbook debacle? MS acted smart allowing OEMs to ship with XP again and all the market they lost to LINUX distributions like UBUNTU was almost totally recovered, without investing 1 penny in R&D because they just used and old product that they knew will move the netbooks well. After all “weak” netbooks today are way more powerful than the average desktop 10 years ago so…
Apple isn’t going to take that market share away from them, Apple knows it, that is why they have shifted to this super closed hardware/software ecosystem aimed to people with medium-high to high level of income, which mean never getting over the 20% market share in any field (OS, phones, etc) But they don’t care, because it makes them a lot of money anyway.
And back to Linux distros, the problem is not the system, and is not the unwillingness of people to change, IT IS THE APPS STUPID!!! (no that you are stupid, I am just appropriating an old joke). Only when developers start to create apps for linux (COMERCIAL APPS THAT MAKE MONEY) will average Joe start to consider have it on his business and home. Problem is, developers will not create apps for a system that has such a small market share, is a vicious circle.
As SaaS evolves, MS iron grip may loosen, but there is a danger that just like Apple did, MS could (will?) try to develop a close ecosystem to capitalize in their current dominance, which would trap us even more in this situation.
Solution? Grab the market share by cheating. Forfeit this nonsense about not allowing proprietary code on your free systems, keep the forks, but make it sure that they are still compatible. And more important: Develop a true transparent sandboxed windows/DOS emulator capable of running any windows game or general app on Linux, as a integral part of the system, not a program you have to install or God´s forbids “configure”, which is a geek code word for: “it aint gonna do shit until you learn how to make it do it”.
By being able to run all the old and new apps and games, and that ONE app that keeps Joe under the tyranny of MS. And when a decent market share is in Linux hands (maybe something like 15-25%, but a % of distros that can all work together, it is no use if we have 20% but divided in 1% each that are not compatible with each other) THEN AND ONLY THEM may developers listen and start making native apps for Linux, because EVERYONE FOLLOWS THE MONEY.
Heck if this succeeds even MS will star making stuff for Linux, they may even launch their own fork.
Then you will see true innovation again… because EVERYONE FOLLOWS THE MONEY.