So.... I mean... strictly according to the letter of the vexercise you're not supposed to edit anything visual from that first video to the second, but like with any of the vexercises, it's also your creative work, and you should do what feels right; I think this one can be especially hard when the timing of the text is so attuned to the first song and it doesn't line up with the second. You are not alone with finding that to be a sticking point, and it would be understandable if you went in there and fiddled!