Apple has done this (multi-camera/spatial fusion) since 2016, in addition to multi-frame/temporal fusion. The OS decides whether these techniques are worthwhile for any given shot based on context.
https://developer.apple.com/documentation/avfoundation/avcap...