Notice that here there is no physical superimposition of the advertiser's content in the status bar on the underlying content on the main screen. Does that make a legally important difference? In ordinary framing no such superimposition occurs, either.

Is the combination of audio with a silent film the preparation of a derivative work? Colorization almost certainly is. What's the difference? Does it matter just what the audio is? (Music, words, the same words as the captions....)