The point of a firewall is that it creates a boundary between internal and external. Traffic going through the firewall passes from internal to external. So for all intents and purposes, consider the firewall to be wholly internal except for the 2-dimensional plane that is public-facing; in other words, the network is internal including the firewall up to and excluding the firewall’s doors.
As a result, everything authoritative the firewall does with regard to the network (and again, I’m simplifying) is within that internal boundary. And we can assume the internal network operates at a high trust level such that privileged entities within the boundary (such as a firewall) can see and control certain aspects of traffic without breaking CIA.
The threat actors you’re describing here (governments and so on) are external. Outside of the network, nothing can be more privileged than anything else. So to answer your question, as far as at least the civilian population is aware, no, it’s not possible for external parties to interact with public traffic as though it were a firewall.
Does that make sense? Comment if you have any other questions. I think there might be some core concepts here that would be good to research and review, including network fundamentals and how HTTPS works.