I believe rather my own eyes over a long career than these surveys. It's certainly well below 10% if you don't count being just used for the lack of any alternative (aka sshing)

Placing greater weight in our own anecdotes over randomized trials and researches is what gave us some of the worst experiences in recent years.

StackExchange surveys are not “randomized trials” and only qualifies as the lowest quality research.

The group of devs makes all the difference.

In my current role (contracting at a gov dept in a part that uses Rails and PostgreSQL) it's about 30% using vim or neovim.

In the rest of the department where it's .net and Microsoft all the way, the devs likely have never heard of vim.

Based on my eyes, viruses do not exist. I can't see them anywhere. A sample size of "what I see" is not the best foundation for any claims.

If only more people knew about tramp-mode.