You're misinterpreting that graph. The key detail to note there is that the bars don't add up to 100%. Meaning that it's showing what percentage of people use that IDE/editor at all. The phrasing, "~25% of web devs... used vim as their editor," implies that 1 in 4 web devs use it as their primary or preferred editor, which is not something that can be inferred from the data.
For my part, when I answered the survey, I checked vscode, intellij, vim, jupyter, emacs, and rstudio. Most of those are there because they're my preferred editor for some task or other, but vim is only there because it's the only editor I can expect to have available on every server I ssh into. So, while I did say that I use vim in the survey, I would encourage you not to count me as a vim user for the purposes of any popularity debates.
For my part, when I answered the survey, I checked vscode, intellij, vim, jupyter, emacs, and rstudio. Most of those are there because they're my preferred editor for some task or other, but vim is only there because it's the only editor I can expect to have available on every server I ssh into. So, while I did say that I use vim in the survey, I would encourage you not to count me as a vim user for the purposes of any popularity debates.