I’m not a sysadmin but
Windows Server is amazingly reliable. But what happens when you get that one weird problem?
I regularly see a story about someone’s epic journey starting with an application level bug and ending with debugging some Linux internals and finally solving the problem (some obscure Intel CPU bug, or driver software issue, or epic network race condition etc). Those journeys begin with the belief that with sufficient motivation you can track down any problem on Linux/BSD.
When you watch someone solve a Windows Server bug the “solutions” are very different, and you rarely hear of someone debugging drivers or OS issues.
When I was smaller I wrote embedded software, and tracked down a very-hard-to-find bug in a RTOS.
Disclaimer: The business I helped found depended on Windows Server, and it rarely let us down.