Not too far off …
In case this is helpful to someone, here’s my follow-up.
First, I was able to acquire a WAV file of the entire interview to work from. (Turns out Instant Seminar keeps a backup.) Then after an hour or 2 so of researching the info offered in this thread and messing with spectrum analysis and EQ, I determined that the differences between the first and second parts were too many and varied (for me) to bring them close enough together so the transition wouldn’t be jarring. So I did what I do best: played it by ear. With lots of back-and-forth experimentation, I made an EQ which I applied to the first part to mitigate the muddiness. Then I cut together a short bit of conversation around the call-dropout-and-reconnect, and added the 2nd part to that. Finally I mixed in the intro and outro and exported to the mp3 required for podcast tagging. Not perfect, but not bad.
That’s when my id3 tag editor refused to read the file, but that’s a story for a different thread…
So thanks again to all of you who offered help. It was all useful info, even if I didn’t utilize it all.
Dan