There are plenty of startups solving hard and complex problems. The issue is that they're usually not amenable to the Silicon Valley VC model of "sell it to the next biggest fool."
I think it's a lack of synergy between the angel model and the price of solving really big or complex problems.
YC in my understanding is catered at companies who are solving natural next steps at the right time, not necessarily problems with a 5-10 year horizon.
I am always reminded how Interval Research went from 10 years research to market horizon to suddenly 3-4 years.
For all the Silicon Valley is good it it seems to be best at creating relatively obvious solutions for fairly trivial problems.
If more big and complex problems were solved Elon Musk would be the rule not the exception.
This is what's kind of sad about SV these days.
It might drive the digital economy but one might ask in what direction.