For topic (B) companies can set or not set the header based on GeoIP. Not perfect but GeoIP is already used in load balancers, web servers and applications.
For (B), your proposal requires the website have a database over current rules in every country they would be accessible from. Would a website then, in your opinion, be responsible for the accuracy of this database? We have to presuppose an official GeoIP source that would then be legally binding and under democratic control, but given such a database, would a website serving a wrong header to an IP associated with a specific country then be committing a crime in that country? What would the punishment be?
For (B) this is already a thing. Porn sites and already doing this. Instead of blocking a region I am proposing to stop blocking and instead the law permit them to just add a header. The only people I can imagine apposing this are commercial VPN providers.
I think a better example might be places like polymarket (not allowed to operate in us) or usatoday (serves an EU only version with no cookies). The technical limitations on those systems are both GeoIP as far as I'm aware, and that seem sufficient for regulators.
What I find more interesting is that what you want is within the scope of this law. It's only required that the operating system takes in your age from an admin, from there the application (user agent/webbrowser) is supposed to handle the blocking, which it could do with a header as you suggest.
I will note that you are going to find a lot of libertarians that would oppose banning GeoIP circumvention.
B) How would your RTA header intersect with content rating in different jurisdictions? What if the content is illegal for children in Turkey but legal for children in Kentucky?