"Disruptive" is popular because it implies the product or service can generate value on a large scale without playing catchup to existing market leaders.

If you're going to make an incrementally improvement, you have to first catch up to the pack, and in an established industry, this can be (or at least seem) cost-prohibitive and risky. "Disruptive" startups, in theory, do an end-run around the barriers to entry.

The article suggests that we replace disruptiveness with usefulness as a criteria for evaluating startups. I think we need both; disruptiveness ensures the startup has a fighting chance against the incumbents, and usefulness ensures the startup has a market.

Incidentally, several of the author's supporting arguments can be applied to the term "useful" as well. "Most technology we now consider 'useful' wasn't conceived that way", "the creators of useful technology often don't make the money".