Summary: Ready, with one outstanding question

This document is good at describing its extremely limited scope of restriction - there are a hundred ways to slice "problematic characters in Unicode", including such things as overlong combining character sequences, illegal sequences of directional controls (unpaired U+202C, for instance) and so on. 2.2 does a reasonable job of describing the limited role of this document, and to warn about what other kinds of problems may exist.

One class that I'm surprised to not see mentioned is "Private-use" (https://en.wikipedia.org/wiki/Private_Use_Areas). While it may be inappropriate to limit their use by these profiles, it might be worth mentioning together with the PRECIS reference in 2.2 as a class that protocol designers can think of avoiding. (The "private" blocks are marked DISALLOWED in the PRECIS registry at https://www.iana.org/assignments/precis-tables-6.3.0/precis-tables-6.3.0.xhtml?&page=11 - but that's a few steps to follow).

Apart from that, I think this document is in shape for publication.