Universal-character-names are in phase 1, before we can figure out
validity. That is, if you have a bare surrogate, it may be ill-formed, but
we find that out after phase 1.
It's also not clear, for example, that it would be ill formed to have a
string with surrogate pairs in it expressed in Uxxxx notation.
At least reading quickly.

On Sat, Feb 29, 2020, 09:46 Lyberta via SG16 <sg16_at_[hidden]> wrote:

> Tom Honermann via SG16:
> > we had general consensus (we didn't poll) for renaming
> /universal-character-name/ to /unicode-code-point/
>
> As usual, if we don't allow surrogates there, I strongly propose
> /unicode-scalar-value/ instead.
>
> --
> SG16 mailing list
> SG16_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/sg16
>