Earlier quoted context omitted.
Perhaps part of the problem is that we incentivize high-risk ideas too much.
I think it's the opposite. I'm not talking about "Let's start a social media app for Aardvarks!" kind of ideas, but actual scientific research. We used to have pure research orgs like Bell Labs, but companies aren't funding pure research any more, so we need more investment from government. Not every idea should have a path to monetization before it's funded.
I don't equate "skunkworks" with high risk. To me, that's not "risk" as much as "I don't care about monetization in the first place", like you're saying. I agree that research without promise of a payout is sorely missing these days.
When I say "maybe we should have less high risk", I'm thinking of fly-by-night companies that do make a social media app for aardvarks, spend a bunch of money on slick marketing so they can pump their IPO, and then pull the rug with them when they bail.
But billionaires don't become billionaires by doing skunkworks projects. They often do the latter.