This is an unpopular position, but then again, I’m kind of known for them.

MovieTalk is not, IMHO, properly part of TPRS. It’s fine to do if it’s done properly, but it is not TPRS and it, again IMHO, should not be taken as a replacement for TPRS. One more IMHO for good measure: MovieTalk should be used as a dessert, not as the main course of, well, the course. 😉  And the chef (teacher) should know how to cook main courses (do classical TPRS). A pastry-chef only approach is not going to make a successful restaurant.
The reason is this. TPRS works because of comprehensible input, but the reason that works is that kids are listening. They listen because the topics of TPRS are either the kids themselves or things the kids are proposing to talk about. They have a very high degree of creator ownership in the stories asked or read in class.
MovieTalk by its definition depends on a preset storyline (unless you turn it off suddenly and ask kids to tell the ending, which moves you back into TPRS really). The kids had no voice in creating that, and it is not customized to them in any way. A skilled MovieTalk teacher can certainly elicit details and impose them on that storyline or characters, like having the kids name them and so on, but the visual is still set. There’s no way for the class to “decide” what the character is wearing — he is wearing what’s right in front of them. Ditto the locations (unless it’s a very visually-open-ended thing like a minimalist animation), the actions, and so on. MovieTalk is destined because of these things to be at a lower level of personalization or customization than story-asking.
But MovieTalk is easier than classical TPRS because it doesn’t require the creation of a storyline in cooperation with a classful of kids. It doesn’t require as much personalization or customization (though as said it’s better with those things going on). The circling questions can be entirely pre-planned becuase the story will NOT deviate any more than a train will start running eastward if it’s on a set of tracks going north. But the “hook” is the flashy video (flashy in the sense that it’s a tech toy kind of thing, I mean) with the animation or the effect or the music or whatever things are there. That scares me. I don’t want my hook to be external, ever. I want my hook into those kids’ heads to be the heads themselves — the things that kids truly care about. The lifespan of a cute video as an attention-grabber and more importantly an attention-keeper may be longer or shorter depending on a whole bunch of things that I probably don’t even know about
Sure, videos are things that kids like. Their topics, and their cute design, and all that, those are also things that kids like. But the whole point and power of TPRS is that there’s nothing, NOTHING, that kids like more than….themselves. If teachers (and many teachers are doing this these days) only do MovieTalk, or do MovieTalk as a replacement for TPRS because TPRS is “hard” or because they haven’t quite “gotten” TPRS yet, they are relying on the drawing power of that video. And one day that is going to fail.
Now, you can achieve personalization by doing parallel discussion with a MovieTalk. But that ends up being…PQA. Which is TPRS.
Don’t forget that kids usually throw the fancy toys aside eventually. The thing they can play with for hours is the boxes. Because it’s about their imagination and creativity and themselves.