Mapping the world from smartphones and chat forums and reddit threads.
Open source science
Without a huge research budget and dozens of managers, you won't be able to coordinate a ton of researchers. But the good news is, you didn't really want to coordinate everyone anyways. You can just open the gates and let people get to work. It works fine for software!
Another paradigm for crowdsourcing is to gamify science. That is not my jam; I feel that participants deserve a better stake than only twiddling their dopamine system and the incentives of recruiting that way are not ideal — scientists’ games will not be as fun as real game company’s highly evolved games, and the participants we can recruit for addictive behaviour are probably invested in a suboptimal manner for our goals.
Software frameworks for data collection: (Thanks Dan Pagendam for the tip.)