They tried that strategy with game console DRM, and it was defeated with several methods, from boot device swapping to invisible hypervisors to override chips.
It also requires that they find someway to deprecate all existing consumer hardware. Too much current inertia vs a swarm of opponents who are too nimble, and not enough time and runway to get it done.
The best approach would be to try and implement gating at the ISP level, but that would require a bunch of extra investment for packet inspection, would drive latency times up through the roof, and would need a way to detect steganography in the general case. That last one is the halting problem all over again.
In fact, the whole task boils down to "don't let those dirty goyims do or say anything we don't want them to do", which is the halting problem all the way down. Censoring language just causes new cyphers to develop, censoring video streams just makes steganographied equivalents pop up. Trying to tie data origination to a state ID schema just causes people to use fake IDs. Fundamentally, this war was won by the fundamental design of the TCP/IP protocol since the beginning. If a nuclear war cannot stop the data flow, a censorship war doesn't stand a chance.