Is it sort of a SpaceX model? Rebalance the relationship between iterations of theorizing and of practice? Unclear how many iterations of practice they'll be funded for.
Of course the payoff is high. In the picks-and-shovels end of the information economy, it's clear that power is a good business.
It seems a bit SpaceX. They are trying for net electricity out in the current iteration. If they get that it'd be a huge deal and funding no problem. Big if though.