The timing of the video capture is almost certainly configurable via software and can be done without rebooting as well as a number of other things related to the video format change, but I suspect the reboot is due to simplifying the software. There’s probably a lot of things done at boot time based on the video format that’s just easier to reboot than to write code that can change those values on the fly. Sure rebooting is an inconvenience to the user and it’d be better to allow it to be changed without rebooting, but probably most users don’t change video format often and requiring the reboot means they can make the software simpler (and more efficient).
I think the bigger question is: why does changing NTSC/PAL require formatting the memory card? The frame rate of the video is independent of how the data is written to storage (aside from analog media). Even if there was a separate NTSC/PAL clock for the video capture side, the storage device doesn’t need to use the same clock and even if you did change the clock of the storage, that wouldn’t change the actual data being written, thus no need to reformat.
The clock signal for storage writes is independent of the data being written and in fact, for all solid state media I can think of, the clock doesn’t even need to be consistent. The clock speed is effectively variable as long as other timing constraints are met (think things like max speed, but when you get in to electronics hardware, we have to get in more detail like how fast does a signal transition and how long from one signal transitioning until another is valid), but again that’s independent of the video format. It’d be like saying you can’t put a PAL and NTSC .mkv file on the same memory card or that you can’t read a memory card in a different device because they were read/written at different speeds. This absolutely can be the case with analog storage, but not solid state digital. At least none that I’m aware of.
The timing of the video capture is almost certainly configurable via software and can be done without rebooting as well as a number of other things related to the video format change, but I suspect the reboot is due to simplifying the software. There’s probably a lot of things done at boot time based on the video format that’s just easier to reboot than to write code that can change those values on the fly. Sure rebooting is an inconvenience to the user and it’d be better to allow it to be changed without rebooting, but probably most users don’t change video format often and requiring the reboot means they can make the software simpler (and more efficient).
I think the bigger question is: why does changing NTSC/PAL require formatting the memory card? The frame rate of the video is independent of how the data is written to storage (aside from analog media). Even if there was a separate NTSC/PAL clock for the video capture side, the storage device doesn’t need to use the same clock and even if you did change the clock of the storage, that wouldn’t change the actual data being written, thus no need to reformat.
The clock signal for storage writes is independent of the data being written and in fact, for all solid state media I can think of, the clock doesn’t even need to be consistent. The clock speed is effectively variable as long as other timing constraints are met (think things like max speed, but when you get in to electronics hardware, we have to get in more detail like how fast does a signal transition and how long from one signal transitioning until another is valid), but again that’s independent of the video format. It’d be like saying you can’t put a PAL and NTSC .mkv file on the same memory card or that you can’t read a memory card in a different device because they were read/written at different speeds. This absolutely can be the case with analog storage, but not solid state digital. At least none that I’m aware of.