I dubbed in the messed-up lines and did an audio compression on chapter 2 of Trang--it went fine (and the compression took care of that pesky clipping all right), but I decided to give it a listen-through. Well, I'm happy I did--there's a character in that chapter who gives a speech early on. He's the only character talking then, so I just did it in my voice, but later on, several characters are talking, so he took on more of a character voice. It was a subtle voice adjustment, so I didn't think it mattered, but it really does: Even a subtle character voice has far more continuity (it's always soft, or it's always deep) than my natural speaking voice. So I re-recorded his speech, compressed it, and edited it in.
I'll do noise removal next. So far, it definitely seems like there's no point in doing noise removal before compression--I haven't see any drawback to waiting, and if you do it before, you just wind up doing it twice.