Why do you think it's for TV coverage? It seems plausible to me it's just better-suited for the medium (of a remote call).

If you listen to the in-person oral arguments, there's a ton of interruption and pre-emption all the time (with the Justices always winning the pre-emption battle against the attorneys), which works fine with the visual cues and 0 latency of being present together. If they tried to imitate that same style over a typical Zoom session it would be a disaster, people talking all over each other, etc.

Actually, the change probably has more to do with the difficulty of social cues for turn-taking over videoconference calls.