Since there are so many users with browsers, I'd be curious if data is available that filters out consumer users/residential IP addresses etc. To better understand business browser usage and get more usable percentage figures.
Wouldn't be surprising to find that security conscious people are more apt to alter their user agent string or have a plugin that rotates it often to counter fingerprinting. There are probably fewer genuine IE11 users than what is being detected.
>or have a plugin that rotates it often to counter fingerprinting
That's not going to fool any competent fingerprinting scripts. Blink/webkit/gecko have different javascript implementations so it's easy to cross-check whether your user agent header matches your "real" user agent. The only thing you're really going to fool are server side logs for user agent. If anything, using user agent spoofer is an entropy source (makes you stick out more) because most people don't spoof their user agent.
Rotating the agent alone, right, but it's part of a larger set of tools usually. I don't think anyone is implying that security conscious users are only doing that.
On the contrary, you could take the WhatsApp stance and say that everyone deserves a more secure email platform, not just those who know enough about technology to know that the default browser is not good enough. Not saying it's an easy task, of course, but that could be a goal for protonmail.
The typical scenario is forced use by enterprises where the employees have no choice. But now that Windows 7 is on its last legs, this may be less of an issue.
You can have IE 11 on Windows 10. It's what I have, since the software we're working on use ActivX (to communicate with card reader software on the client machine); we should finally move to Chrome this year.