Because "making everyone identifiable" isn't an explicit design goal. Rather it is merely an implicit imperative (of Facebook et al, who are pushing these laws) that casts its shadow over the design. That shadow is what results in a design based around sending identifying information from the client to the server. Once this dynamic is normalized, servers will demand ever-more identifying information and evidence that it is correct.

Note that this design does the exact opposite of giving parents control to protect their own children - rather it puts the ultimate decision making ability into the hands of corporate attorneys! For example, we can easily imagine a "Facebook4Kidz" site that does the bare legal minimum to avoid liability for addicting kids to dopamine drips, and no more. Client side software based around RTA headers would allow parents to choose to filter things like that out, whereas when the server is making the decision its anything goes as long as the corporate attorneys have given it the green light.