repeat_* noexcept fixes

[justend29]

My original goal was to conditionally add set_error_t(std::exception_ptr) to the completion signatures of the repeat_* senders, only to realize that exceptions from connect() was one of the locations set_error() could be called, where the nothrow-ness of connect() is only discernible once a receiver is known (after advertising completion_signatures)

Nevertheless, on my travels, I learned a great deal about the meta-programming in this library and made some fixes:

1. repeat state base classes are abstract but were missing virtual destructors -> made them protected
2. repeat state classes were missing virtual destructors -> made the classes final
3. add noexcept specifiers on __connect() and propagate to the operation states

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[RobertLeahy]

where the nothrow-ness of connect() is only discernible once a receiver is known (after advertising completion_signatures)

Note that this is the topic of P3373 which is part of C++26.

You can use ::STDEXEC::__receiver_archetype<Env> to check whether or not connect will throw without a receiver.

[justend29]

You are literally the mystical wizard with all the answers. I'll review the provided info and take another stab at the original goal. Thanks, @RobertLeahy!

[justend29]

Original goal lives on #1799