The minor issue is the quantifying of developers. The methodology for the 15-20 million figure isn’t clear, but it at least passes some basic sniff test.
What’s completely bullshit is using stackoverflow surveys. There is no reason to believe the respondents of that survey are an unbiased cross section of the 20 million and many reasons to believe they aren’t.
I would totally expect the stackoverflow social 2.0 tech hipster webshit crowd that would fill out that dumb survey to be more likely to use Go. The same with oreilly surveys. Old grizzled grey beards aren’t as likely to fill out publisher surveys on old stable tech.
Also, answering that you have “used” go is not the same as being a regular developer (as noted by one of the commenters there... assembly is more “popular” than Go).
OTOH, regarding
> answering that you have “used” go is not the same as being a regular developer
The question they ask in the survey is: "Which of the following programming, scripting, and markup languages have you done extensive development work in over the past year, and which do you want to work in over the next year?", so it's more than just having "used" it.
Also I don’t think anyone that participates on this website has any moral superiority over tech hipster webshit. But it’s good to realize this is a skewed demographic.
This seems a little implausible to me, can this really be the case?
Edit: Oh, I see the reference in the post https://research.swtch.com/gophercount (I'm still skeptical!)