If no one else is preparing this PR I will, as the changes in P2900 depend upon it.

I don't think P2900 merge-conflict-depends on the wording changes here; it's just that P2900 likes to use some words introduced here.

no we also have edits to the new paragraph introduced in intro.abstract about observable checkpoints. I have almost completed a PR for P1494 (and I think it's good practice for me to start with a small paper first :) )

Reassigned.

Maybe you want to base the pull request of P2900 on the one for this paper here, then, instead of on the "main" branch.

Other than how to specify a reference into the C standard, I have the PR ready for P1494. I think we will be able to get that reviewed before i finish the full PR for P2900.

There's no core/library feature-test macro added. Was this intended? (I'll submit issues if not.)

I don't think we need a core feature-test macro. This doesn't add a feature, it constrains implementations around undefined behavior.

Whether LEWG wants to have one for std::observable, I don't know.

I've been informed that LWG didn't discuss it directly but that the LWG chair indicated that he thought this was too obscure and it wasn't worth the bother to add a feature-test macro. CWG did discuss it and, as Jens said, they didn't want one.