Interesting discussion. I think it’s one of those situations (UAS + terrain) where it’s wise to consider what the greater threat is: flying into the ground because you applied the UAS drill or LOC into the ground because you didn’t.
I would be very surprised if 10degs and 80% (not on the 737 now but that seems about right?) gave you a shallower climb gradient than an engine failure at V1 on a full derate, so the getting-away-from-the-ground bit should be pretty much assured. The danger is that if you delay the UAS drill for too long, you may be in a position where it’s not going to work as effectively (or not at all) - I don’t think Boeing guarantee a successful outcome right out at the corners of the flight envelope, which is why prompt diagnosis and action is essential.
I have no insider knowledge but maybe there was concern that applying the drill at or during rotation wasn’t optimal and might delay gear retraction, thereby rendering the default pitch and power for UAS unsuitable. Better to get it away from the ground then go for it once the gear is away and a climb established? That would normally be not far away from AA, so standardise on that?