iOS is very good about sandboxing and only letting apps run things while the app is open and focused on. It shows green and orange dots when the camera or mic is being used, and none of my use them without saying so and they only do so when they actually need them. If that is the case, are there any potential privacy issues with it?
Sounds like you trust iOS sandboxing. I have my suspicions it’s not perfect, so how big of a deal it is depends on your threat model. If I’m on iOS I tend to manually enable the camera if I ever need it, and I also avoid using apps when possible.
If I needed to use an app daily that requires camera access, I might change my habit because that sounds like a lot of work.
I agree. To your last point, I think you need to add more nuance. Lots of apps can take pics and thus require access same as mic. It’s not always a red flag until it is, which naturally should bring you back to a threat model you are comfortable with when analyzing whether permissions are valid