So you're saying that they should analyze both audio and video to increase the quality of the captions, if the video has hard-coded captions? I guess that's pos

So you're saying that they should analyze both audio and video to increase the quality of the captions, if the video has hard-coded captions? I guess that's possible, just a question of effort vs. payoff. Inaccurate auto-captions for videos with hard coded captions probably isn't a big enough pain to warrant big investments?