Personally, I'd take advantage of the ability to choose a profile based on source format, although I suppose bitrate would work, too.
FLAC I always convert to 320 CBR. M4A I convert to VBR 0, but I think they're always 256 VBR to start out with (from iTunes). I know M4A to MP3 is lossy to lossy, but tagging support for M4A is so crap that it's worth it to me.
Since we are freewheeling now:
I struggled a while about what to do with formats other than flac or mp3. (which are my two preferred formats by far)
If it was only for tagging purposes I don't like to have formats such as m4a, wma etc. in my library.
That gave me two options:
1. Preserve the sound quality and encode them to flac. But then I would have flac files that would give the impression they were of lossless quality witch would be false. So that was a no.
2. Re-encode them to mp3. But that would result in an additional loss in quality. Not something to lose too much sleep about, but still not very desirable. (and I might loose my audiophile badge ;-)
Then a light bulb appeared above my head.
Any lossy format other than mp3 gets converted to ape.
Ape will contain the original audio without additional loss, and it presents no problems related to tagging.
So now my library contains only three formats that each present a clear purpose to me:
Lossy that was converted from some other format: ape
Easy to remember, and no tagging challenges anymore.