This is pretty anecdotal. We'd have to come up with ways to quantify it.
The story of the failing company taking a moonshot or making risky bets is pretty common.
This sounds like something that there'd be an HBS logo on, it likely already exists. It's a fairly easy thesis project.
There's an LLM that does conceptual searching of scholarly literature https://consensus.app/search/ but it looks like just a classic IDF/bag of words style approach labeled "ML"