In this case, the video was the minimum viable product. The MVP validated Drew’s leap-of-faith assumption that customers wanted the product he was developing not because they said so in a focus group or because of a hopeful analogy to another business, but because they actually signed up.

I think this is a very interesting quote. For one thing, it urges us to expand our notion of what a product is. It isn't necessarily an actual working piece of software; it can be a mock-up or faked demo -- but in any case it's still something that can be shown to potential customers.

But the question that some of us are raising is, when could the video be called an MVP? The answer we would suggest is, it becomes an MVP only after it receives an enthusiastic response from the market. Note that the Ries quote is completely consistent with this interpretation: by the time Ries wrote it, the video had received such a response. Our argument is that while Drew Houston was making the video, it wasn't an MVP yet: it was only an experiment, an attempt at making an MVP, but not yet a validated MVP.

So I disagree that "MVP" is "a term of art and can't really be broken down word by word". Rather, I think that people haven't understood exactly what Ries and Blank mean by it.