I am not a fan of PIP in this situation, but on the home CCTV i do have several cameras going in one picture.
It is also rare i have used PIP in my youtube uploads of traffic, or comparing 2 dash cameras, cuz comparing you still need the 2 pictures to be the same time, so i grab the middle 50% of each and then put them on top of each other to form out 1 video.
Though i guess you could argue it is PIP just both pictures are the same size, i am more inclined to call the traditional Picture in picture ( one big picture + a small one on top of that somewhere ) for picture on picture, but thats probably a grammar error on my Danish part.
To my strangely operating brain picture in picture mean that if you remove the little picture there would be nothing behind it, where as picture on picture, if you remove the little one the parts of the bigger one would still be there behind it.
Also not a fan of the map being in the footage, not in videos, or just reviewing footage on the computer, hence why i do not use dedicated dashcam player software's in general.
I would use that if i got some mystery footage and i would need to know where it is filmed, but there are no mystery to my own footage.
For some select kinds of action footage i do some times like overlaid metrics on the footage, like you see some people do with action camera footage, but it still have to be moderate, for instance i would not like to watch someones 20 minutes long descent of a mountain on a MTB to display every little twist and turn the whole way down.