I think of more physical projects like RepRap (reprap.org) OpenFarmTech and SpaceX that are really pushing things in a new way. I'm sure there's plenty more if people look for them outside the software space.
Web applications don't seem trivial -- they are trivial. Scaling them means being creative, but don't confuse that with true innovation.
I'd argue that innovation happens in widespread tinkering. So it could be true that most of the tinkering today happens to be on webapps. But it's presumptuous to assume that none of these attempts would result in innovations of significant worth to humanity.
We'll remember Facebook in ten years the way we remember MySpace now: We don't. Facebook changes nothing.
If you're working in the same space as Facebook, you're not doing yourself any favors by denying its impact. One, it harms your perspective, and two, it harms your credibility.
Cassandra came from Dynamo; And Dynamo is from Amazons R and D efforts. Btw Amazon is an innovation as it leads to large amount of savings in energy.
Thrift,Tornado,HipHop -> not innovations useful programming tools but not gonna change anything.
Rarely any innovations have come out of Web Applications. In fact the best innovation in data mining [The correct term for Big Data], have come from Universities or R&D efforts of IBM/MSFT/Google/IBM/Amazon.
“The best minds of my generation are thinking about how to make people click ads. That sucks.” Sound true at face-value. But guess what? In order to process the huge amounts of data necessary to 'make people click ads', we use map/reduce, and we get and/or improve Hadoop. Both of which may then be used for other Big Data applications.
Sure, web applications seem trivial. But the innovations created as an aside to them very often are not.