> part of the Unicode Category Po. So it shouldn't interfere if we
> consider the Unicode categories Ps (resp. Pe) as valid context for
> opening (resp. closing) constrained markup. Or am I wrong here?
As I stated above, it's outside the scope of this project to redefine AsciiDoc. AsciiDoc as we know it today uses ASCII for markup characters. (That's where it gets its name, as I have always understood it). It's not appropriate at this time to change or expand the characters used for markup. That can be a conversation for after 1.0 is finalized, and even then we'd have to give it considerable thought and discussion. But we have so far to go before we get there that I just don't see the use in spending time on it right now.
The discussion about space separators is different since space separators are both user content and markup characters, so we can entertain some more flexibility there to deal with edge cases such as a non-breaking space before an end punctuation (e.g., full stop question mark).