I wonder what do you think are the real benefits from doing "Discovery" questions?
TL;DR: Discovery = What do we make? Usability = Did we make it right?
Usability, as the term is now used, is evaluative. It's about figuring out whether a particular web site / app / whatever is usable. Discovery, in the way that GV are using the term, is generative. It's about discovering the needs and pain points of a group so we can figure out what we need to build.
Both evaluative and generative user research are hugely valuable - but in very different ways.
Generative / discover related research is all about figuring out what problems to solve. Evaluative related research is all about figuring out whether our solution actually solves the problem.
For example we're in the middle of rebuilding our company website at the moment.
We'll be doing usability testing to figure out whether folk can navigate it reasonably, whether they can find the workshop descriptions, whether the can contact us, etc.
However, the discovery work that have already happened has told us about the pain points our customers have, the kind of language our customers use, how particular terms are ambiguous, etc. Which in turn has led us to develop and market specific workshops with specific language, etc.
Usability testing reveal very little about whether we have the right workshops or not. Discover would reveal very little about whether people can book a workshop.