All this while open source projects would be the perfect place to really shine with innovation and with something new. Why just copy, why not create something new?
First of all, I don't think few things are really new in terms of technological development. Most technology, if not all, is built upon a layer of old technology. The wheel was probably an evolution of the rolling log, a technology that nature invented. Famous computing technology examples: the Macintosh, which "innovated" on top of technologies developed at Xerox's Palo Alto Research Center, and Windows, which ... (you get the point).
But if by "new", you mean fairly new, then it's easy to trace the apparent lack of innovation to the shortage or more often absence of designers in open source projects. (This is something Canonical has been trying to address with projects like Unity with I'm not sure what level of success.) Designers are a necessity when dealing with graphical programs which are spatial, as against the more sequential nature of programming.
You can find lots of innovation on the console, which includes software like emacs (an innovation against the line editors of the time), the object-oriented scripting language Python, and even Unix itself. These are "old" examples of innovation, which I mention only because they are pretty well known.
On top of that the shooter games are all based on ID's engines that are many years old.
Not all open-source games are based on the ID engine. Here's one that isn't.