As a background the C++ committee decided not to merge the network TS (i.e. ASIO) and instead focus on the sender/receiver proposal. This has been contentious to say the least as ASIO is industry proven while the sender/receiver is more greenfield.

Sender/receiver is also quite complex (although ASIO is not exactly simple) and most likely not ready for C++23, which means that C++ won't have networking for at least 6 more years.

Personally I don't have a strong opinion either way, I like both ASIO and S/R (and have reimplemented significant chunks of both). I think the proposals are not mutually exclusive: ASIO should have been merged now and S/R added at a later time when ready. S/R will need a scheduler anyway and might as well be ASIO.