Years of interviewing people for my own teams has shown me that understanding your tools (especially to the point of being opinionated about them) is probably the #1 indicator of success.
Where I work we use a lot of jetbrains IDE's and I've been a massive fan of Intellij, I absolutely love it when I can show someone something that makes their life easier.
Whether it's how to implement an external program properly using all the expansions or how to use bookmarks (how people can use an IDE for years and not find that is a good question).
I've always RTFM of whatever tool I'm using as the first step then I go look at the default keybindings to see what else there is.
The one that blows people away though is the custom exporters you can write for data views, oh you want that result set in a PHP array for a test set, yeah Copy as PHP has you.
Intellij is absurdly powerful without writing custom plugins (which is my next step I guess).
IdeaVim is the uncanny value of vim for me. It's so close to being actual vim, but every now and then it does something unexpected and it completely throws me off.
I recorded a macro to change shouldBe([true|false]) to shouldBe[True|False](). Pretty simple: qqf(ds(~$a()<esc>q
But the vim-surround plugin emulation doesn't work properly with macros, so it'd get stuck and instead of just doing n@@ , I'd have to do n@@( , i.e pass the seperator I used for the delete surround command.
It doesn't have fold toggling, so z[a|A] doesn't work. The undo behaves weird, you usually have to do it twice UNLESS you've not moved the cursor since doing the last edit.
The git integration isn't as good as vim-fugitive and I constantly find my self reaching for the mouse to do anything git related, whereas in vim I could do it all with the keyboard.
The plugin gets brought up all the time when people talk about the JetBrains tools and no one seems to mention the short comings.
(intellij sometimes has recurring problems with lost keyboard input under linux+gnome. IdeaVIM is one of the first plugins i disable, when that happens. and it often helps.)
It's also what's so hard about hiring junior candidates and why so many companies shy away from it.
Truly junior hires don't really know the tools yet. You can't really afford to wait for the rare junior who does, so unfortunately it's not as useful a test in that situation.
That's why I think OP's "seasoned sysadmin" line is perfectly valid. The junior that knows their tools has already cleared the hardest hurdle for most companies. Everything else is trainable through mentorship, but tools usage requires experience.
Indeed, I always advise people who want to be developers to build some stuff to show, to separate yourself from everyone else, and make it easier for the hirer. That is another reason to do it, an IT degree doesn't mean you can write code, as you say the rest can be learnt.
I don't know about that, my first tool was vi, I have strong opinions about it :-). I use an all singing IDE (Appcode) now, its not perfect either but works for now.
Years of interviewing people for my own teams has shown me that understanding your tools (especially to the point of being opinionated about them) is probably the #1 indicator of success.