Nowadays, NTSC isn't really all that relevant anymore since Japan ended analogue television almost a decade ago. But for most of the history of the anime industry, that's how it was broadcast.

Then it strikes me as really odd that the film was shoot at 24fps. To prepare 24fps video for NTSC broadcast, it would have to go through 3:2 pull down, which results in both interlaced frames and changes in frame timing.

Why was it not produced at 30fps instead, which fits much better with NTSC? Surely all that interlacing can't be good?

The basic animation modes should stay the same, shooting on twos (12 cels per second) is easily handled as shooting on twos (15cps) or threes (10cps) at 30fps, and shooting on threes (8cps) is easily either threes (10cps) or fours (7.5cps)

What's the technical reason for picking an animation frame rate that doesn't match the televison frame rate?