I used to run a chatbot that screened people. We thought we were asking simple enough questions, and yet two million chats in, we still discovered at least a Dozen unexpected/misdetected intents that required care. A day. And here is the problem: management will ask: "so if we add handling for 'rotate the map 30 deg', how many people will appreciate it?". The answer is usually "one person every 400k sessions". For an MbA trained manager that means there is no reason to do it. And its the same with every other similar discovery.
Running an NLP based tool requires staffing a 10-20 people full time, telling them to keep the quality high and leaving them the fck alone. Thats anathema for most companies that proclaim to be "metric based". Google is a prime example of this behavior.
But you control that to some extent. If more people are aware of the command and believe it will work (I wouldn't) then you'll probably get higher incidence rates. I wouldn't believe it would work because like the person two parent comments above, less than half the novel crap I try works. So I try novel crap less and less frequently compared to things that I know will work.
I've been wondering why we haven't really had any advances in our voice commands for nearly a decade - things like zooming and such are actions I've long wanted to be able to perform.
Even that rotate map.. when your car is shown at the bottom and map is hard locked to current heading, when navigation is about to do a sharp right you end up with zero visibility of the rest of the route as they still haven't bothered doing a projection and modifying rotation dynamically - they're not even trying.
All your examples are exactly the kinds of things that an intelligent system should be able to respond to, in order to be useful.
Getting that haircut ordered is the _last_ thing we need.