> because the browser has zero idea where these bytes can go after they hit the socket

The attestation uses a secure enclave in your processor with a secret key you can't access to verify that secure boot is on, you booted a signed OS, the OS is in locked-down mode, etc.

>The attestation uses

>you can't access

Don't you see how contradictory this is?

No secure enclave of registers or hidden secret keys can help, because a person can utilize the lower-level physical world around the processor to manipulate it (e.g sending electrical currents from a programator device manually). But that is a last resort, there are simple software attacks available already to fake as many "attested" devices as needed (for the same DRM system of Android). It will only bring more jeopardy to the "integrity"

See that's exactly the issue why I hate this. You can always circumvent it, worst case with an electron microscope and some acid. So all it really does is prevent the average user from gaining control over their own hardware.

And for tech-minded people it doesn't fundamentally change anything, it just means that it now takes more time to do the same than before

True, a cat-and-mouse game going on forever. Anyways, I don't believe they can succeed in walling such a monstrosity of technologies as the web, just by controlling some parts of it, even significant parts like the browser or search. It is only something governments can do by requiring a passport scan each time you open a connection (which is closing when you eject the passport from the scanner)

This is why Risc-V being developed in China and other countries and exported elsewhere is ironically a good thing at the base-level of computing. The chinese computers will require China's bugs, whereas exported good will NOT have it, otherwise it won't be bought.