Some people I guess are just so bright that they don't need it. That is not most people.
Someone contributing to an official Stack Overflow blog seems like a likely candidate!
Sure, the conclusions are the result of a fishing expedition into the usage data for something to comment on, rather than something scientific that began with a question or tests a hypothesis.
However, while the total number of questions isn't so meaningful, that is the denominator used to normalize the data across languages. Think of it as `n`, where the confidence interval for Java is probably smaller than the one for Sharepoint.
I think it is best to think of this as exploratory analysis rather than a study exploring a specific question. It would be interesting to see further analysis into what puts Paris and London on one end and Moscow and Tokyo on another.
Let's say that Stack Overflow makes programmers more productive, but some cultures might have a tendency to think that it is "cheating" to use the site while at work. Addressing that would be an easy way to increase market penetration. Maybe they should get some submarine articles published in Japanese and Filipino trade publications.
I rarely use stackoverflow (like once a few months), how do we factor in the data which we don't have in it? What's the accuracy of the current data? And how can we foem conclusion?
The post should say, "What tags are used to ask questions on stackoverflow at night?" Because, as powerful as they are, SO can't quantitatively decide which languages are used at what time, they don't have the data!! I learned an entire language without ever looking at a SO qn, it was Go language.