This is a pretty broad statement and I find myself in violent disagreement.
Whenever I install a Linux distribution to play with… which is usually how it always ends up, since I run into issues… I find myself installing and using software that I wish would be running on Windows or the Mac. I often find myself saying… “Self… why doesn’t the (such and such) platform have an app like this?”
Linux does innovate – quite well, actually.
The problem is that Linux innovates but cares very little about making software problem free. I often feel that open source developers are pushed by the hunger of making an application bend to their will. They come up with a fantastic idea, write it, publish it and it gets widely adopted by whatever package management application will pick it up. The software becomes popular.
Then, a major bug is found in the functionality of the program… i.e. functionality that is advertised but doesn’t work. Take, for example, many builds of the KMail PIM suite. It advertises that it works with Exchange via plugins but I challenge you to make it work reliably. File a bug against it and what happens? Your bug is largely ignored because the developer or package maintainer does not have a vested interest in fixing it.
In my mind, the problem with open source development today is not that there’s a lack of innovation. There’s just a lack of making it work as advertised.
That will keep Linux in the shadows. That’s probably good for most die-hard Linux fans though.
One other item I will share. A lot of times when I wish that a Linux app exists on another platform – I often find it on another platform… namely, OS X. To top it off, usually the app is done well and relatively bug free as a bonus.