I haven't read the underlying paper, so this probably doesn't detract from the larger point. However, in terms of the Economist's article, I'm not sure that this is a good "sign" of any trend; rather, it would seem inextricably linked to the 2008 crash that led to the Great Recession.
Tech startups before 2008 generally made money by suddenly providing new and cool stuff. Think Google, Facebook, Twitter, Apple.
Tech startups now seem to be focussed to a greater degree on rent-seeking business models- think Uber, AirBnb, UpWork, Stripe, various ginormous tech mega-consultancies.
Do you see the difference? The former have won market share by making products that are clearly better than the competition. The latter are essentially attempts to capture "middle man" monopolies on existing transaction chains.
This is of course a bit of a generalisation, but rent-seeking _is_ a bad thing in a number of non-obvious ways, and should be called out to a much greater degree.
Rent seeking specifically means to extract extra money from an existing asset, but not produce any extra value (but purely by virtue of owning the asset).
You could say that anything to do with the "Sharing" or "Gig" economy is a play on rent extraction. Companies seek to own a marketplace in order to levy a rent on transactions.
I'm not convinced that the more recent rent-seeking companies are much different than the previous generation. Both were seeking to displace the incumbent rent seekers (Uber == taxis, AirBnB == hotels, Stripe == traditional credit card processors) by finding a more efficient or more modern way of accomplishing the existing use case. The only real difference between them and Google/FB/Twitter is that their products are aimed at consumers instead of advertisers.
I also think that even if an entrepreneurial activity isn't "productive" in an economic sense, it could still be useful from the perspective of a person getting out of an employment situation that makes them really unhappy. People starting and failing in a handful of businesses on their way to figuring it out probably doesn't measure out great in the stats, but it's good (hopefully) from a making people happy perspective.
With how much money is being thrown around in salaries and liquid RSU's these days, the costs of creating or joining a startup have never seemed higher (granted I'm relatively new to the industry).