I wonder about this with open-source as well, since it's often about scratching the itches of the demographic that participates in it. But I don't know the unexotic underclass well enough to know if there's anything helpful software could do.
This is why I believe open source will always be much better at creating compilers and programming editors than desktop environments and word processors. There's nothing wrong with this, but it kind of explains why the Year of the Linux Desktop has yet to arrive. It's a lot harder to solve other people's problems than your own.
That being said, I agree with you. Anecdotally, I haven't contributed anything to non-technical open-source projects outside of the odd bug report. That may be a good belated New Year's resolution.
I'm also hoping that interesting experiments in lowering the unnecessary barriers to doing powerful things in programming will inspire folks like me to find a useful way to contribute to one of those summits. We need infinite Alan Kays and Bret Victors.