If so, it just sounds like they're doing an experiment to see if hiring more people but working them less produces better results (Hint, it does in non-dysfunctional workplaces.)
Actually, at least to a point, hiring the same number of people but working them less produces better results.
That's how we got the 40-hour work-week to begin with. It's generally assumed that time working has decreased over the centuries, but that isn't quite true. Medieval farmers, laborers, and craftsmen did work long days (perhaps 9-12 hours), but winter conditions and lack of light with short days meant that these long days were only for short segments of the year. Yes, during planting and harvest, the farmers might work like crazy, but then they'd have a long winter of time to recuperate. This combination works well both physically and mentally, which is the reason studies tend to show that people who never take "vacations" (especially extended ones) tend to be less productive than those who do.
It wasn't until the Industrial Revolution and the migration of poor laborers to big factories, along with advances in tech, power, etc. that workers could be exploited with long hours essentially year-round. The average medieval or renaissance peasant or laborer probably worked around the number of hours a 40-hour/week worker works today. But by the 19th century, factory workers dramatically increased that -- often working 70+ hours most weeks, sometimes with 14-16 hour days. Factory owners mistakenly thought that working their laborers to death (often quite literally) would maximize profit. What instead happened was increased accidents, along with unhappy exhausted workers who would fall ill and need to be replaced with other untrained laborers. (Reforms (sometimes violent) eventually brought limitations down to 12 or even 10-hour days in some places during the 19th century. Unions fought a piecemeal battle to try to get the requirements lower.)
But the largest reform happened in the early 20th century, when Henry Ford actually experimented with shorter work-weeks (i.e., our standard 5-day, 40-hour week) and realized it (1) increased productivity (not just productivity per hour but productivity per worker), (2) decreased accidents and errors (which were a major cause of decreased productivity on assembly lines, since a major accident could shut down the line for a long time), and (3) increased retention for trained, skilled workers, and (4) also had the side benefit of increasing worker happiness. In many cases, the actual weekly output of the same amount of workers who decreased hours from 60 to 40 per week increased by 50%.
Most of the classic studies of productivity have been done on laborers, and they have generally shown productivity is maximized somewhere between 40 and 50 hours per week. But that's laborers, and those classic studies have been undermined by subsequent studies in Europe in the past couple decades which seem to show people doing even fewer (30-35 hours/week) often are more productive than the classic 40-50 hour folks. Also, the summary mentions "engineers and tech staff," whose "labor" is primarily mental. Productivity studies are harder to design for those sorts of jobs, but it wouldn't surprise me at all to discover that for some jobs the maximum productivity occurs at quite a bit lower than 40 hours/week.
Here's the difference today, though -- Ford paid his workers well, in fact increasing his salaries when he decreased the hours, because he saw the productivity increases. His workers responded well and did better work, because they likely remembered grandpa coming home exhausted from the mines and dying from black lung at age 55 -- and the 40/hour week with decent salary was amazing. Fast forward 80-90 years, though, and executives are all about cutting salaries as much as possible, viewing workers as completely expendable, and our cut-throat consumerist culture has taught us that we "need" all sorts of "stuff" so we have to work harder and harder, with longer hours to keep up with the Joneses. Bye bye, 40-hour week... hello, 19th-century factory culture recreated for Amazon tech workers burning out doing 60+ hours/week.
Bottom line: At least in some cases, a policy like this might not only increase efficiency per hour but even lead to increased productivity per worker, even with fewer working hours. But a lot of that will probably depend on work culture, organization, and expectations. If the Amazon workers doing only 30 hours/week are treated exactly the same and given organized tasks in the same way the 60 hours/week folks are, there may be little productivity difference.
And thus the blanket salary decrease will ultimately undermine this, if it's actually an experiment. But it's really difficult to figure out a better way to run it -- because if you start providing salary incentives to people who "only work 30 hours" but are more productive, you'll end up with people punching in/out to look like they're only doing 30 hours, but ultimately are working more hours "off the clock." In our current culture, I'm not sure how to fix this, but I suppose it is at least be nice to see a large company trying to be more flexible.