The value of the data can be as simple as knowing what negative messages/videos you'll continually watch and how those videos, in part, impact larger swaths of the population. And with data sharing between apps and devices, you could likely track actions across much of the web to the influence of these videos.
It's in China's interest, and others, to destabilize the populations of it's enemies. Traditionally this could be through propaganda, but a potentially more powerful force could be to make the population depressed, angry, confused, and distracted. This impacts the real world in your day to day, the communities you live in, and the quality/intentions of elected officials.
Now this could easily be accomplished even with the best intentions from the company creating the software. So how nefarious or not this actually is, or how much manipulation the Chinese government is applying here is totally debatable.
I'm not arguing for one side of that or the other. My point is data is super powerful, is often tied to apps and devices you don't realize, and can be used to motivate action at large scale, so the holder of the data really does matter.
That said, US companies and others outside China also have tons of data and use it for their own profit and potentially to spread propaganda and destabilizing information. In the US, however, the chance of malintent being the driving force is less probable and, as far as I can see, tend more to be influential actors or groups. In these cases, the software company and regulators are more likely to to moderate or suppress this because it can be bad for business. When it's a geopolitical foe who is or simply can cause issues this way, the intentions and lack of regulation are far more problematic.
