Innovation is the point of software freedom
Well, let's look at who judges our software:
Who will judge if we have developed software that we want? We will. We use our software and it either does what we want, or we change it or file bug reports (or complain on slashdot and other forums we created).
Who will judge if we have developed innovative software? The patent office. Or, if the patent office ignores software, then no one will judge if our software is innovative. If innovation was important to us, and no one was examining our software for innovation, then FSFE would have to hire someone to monitor how innovative free software is. But they won't, because it's not important. Clearly innovation was never the point.
Put another way: Are we happy when we have innovative software that doesn't do what we want? No. Are we happy when we have non-innovative software that does do what we want? Yes. (Example: a free clone of an existing proprietary package.)
All things being equal, scientific progress is good, and it will happen, but it's a side effect, not our primary goal.