(now that i have a few more minutes)
the problem with this, in your singer-needs-to-warm-up scenario, is this:
let's say you like take 8 of verse 2 and take 1 of verse 1. further, let's say that over 8 takes, your singer is now singing 5 dB louder on average.
using a compressor, you're now compressing the entirety of verse 2 an average of 5 dB. that compression will impart a sound on the verse, and though your 2 verses may (or may not) be about the same volume, they're going to sound different.
ergo, you shouldn't be using a compressor to do that. what you should be doing is automating volume curves and raising the level of verse 1 5dB or so, so their levels and sound match. THEN you can put a compressor on the track to smooth some peaks, and it'll be doing about the same amount of work across both verses, which results in a closer match of sounds.
***Applauds Zimv20's patience***