The Gallup study referenced in the NYT article you cited says:
State of the American Workplace delivers unprecedented analytics and advice on the changing workplace, using data collected from more than 195,600 U.S. employees, more than 31 million respondents, and insight from advising leading Fortune 1000 companies.
What it doesn’t say is that this is a random sample of all employees. Indeed one of the stated purposes of the study is
enable people to work successfully from locations besides the office
I don’t know the percentage of all workers that work in an office but I’m guessing farm laborers, miners, cleaning staff, delivery drivers, retail workers, etc. make up a significant portion of the working population and I gather from Gallup’s writing about its study is that it is focused on big companies with large numbers of office workers.
One must also take into account the Murray-Gellmann Amnesia Effect. A NYT article is not on par with a study by an expert.
Regardless of all this you miss the point of my original response. You’ve come up with possible objections to a study by an expert. Rather than assume you are the one missing information you assume the expert is being deliberately misleading. A reasonable person would not assume this. A reasonable person would assume that the professor, while possibly wrong, still knows more about the issue and use the opportunity to learn more.
The most natural question by you is, “Why does the professor make the conclusion he makes despite these objections?” One answer is bias. Another is because the objections are easily debunked. Which one do you think is more likely?