It's 15 companies that adopted a technique and it's part of a broader constellation of experiments in this area which have been confounding the traditional logic that productivity would decrease. https://www.4dayweek.com/research

The people representing the businesses in the preprint hold titles like co-founder, CEO/founder, COO, General Manager, and CEO. The size of the business and sector are also noted. I think your framing of them as "unidentified people" is therefore off, it is certainly not the same as a journalist conveniently using "unnamed sources", this is standard academic practice.

Different companies measure different things, but they do measure and that is addressed in the paper. "revenue (DM10), profit (DM4), other financial targets (DM2, DM6), customer/client satisfaction ( DM8, DM6), story points (DM14), sprint goals (DM7), billable hours (DM12), capacity ( DM4), response rates (DM10), standard operating procedure metrics (DM9), sick leave (DM1, D M 4. DM9, DM15), lodgements (DM12), employee happiness (DM6, DM15), projects delivered on time (DM15), and net promoter score ( DM4)". There were also other benefits like hiring and retention.

So this is not what "unidentified people" "thought" about productivity, this was founders and the c-suite using their existing favoured metrics. On those metrics a large number of them reported an increase in productivity, and a larger group reported no deleterious effects on productivity. This is broadly consistent with the trends in the wider research into this area globally, which continually go against the predictions that productivity will drop. Is it universally applicable? I don't think anyone is claiming that.

I've followed this area for a while and, sorry to be impolite, it is your summary that is less accurate than the the one you accuse of being AI-generated.