Poptart's idea sounded good. Why only theorize when you can experiment? Here are the results of encoding "No Such Thing" by John Mayer in AAC (MP4, via quicktime) and MP3 (via iTunes) both at 64kbps:
The numbers are slightly off, but you have to remember that the file doesn't only hold music, it holds track names and a few other very small things (like the extra baggage the AAC (MP4) file had by being wrapped up inside what could potentially be a movie file). And still the files were within about 2% of each other's sizes. The filesize at a given bitrate is the same no matter what the codec.
I tried the same experiment on my brother's PC with Real's RealJukebox using their mp3 encoder and their proprietary encoder and then with Windows Media Player using Microsofts proprietary encoder. All 3 were done at 64kbps per second and all 3 were the same file size (approximately) as the 2 files I got on my mac. Crossplatform, crossplayer, crosscodec -- a given bitrate yields the same filesize every time.
After listening to the MP3 and AAC files with a pair of Sennheiser HD590 headphones, I can definitely say that the AAC file sounded much better than the MP3 file. Very nearly (but not quite) as good as a 128kbps MP3 I made of the same song.
AAC will save you space because you will be able to record at a lower bitrate, not because the bitrates will yield smaller file sizes. The bitrate in the music-ripping world is defined as the rate at which bits are stored on your harddrive -- the bits of harddrive space per second --
after compression, like ladavacm said. If 'ogg' defines bitrate differently then I have no idea how they are defining it and it seems like that would be a rather confusing thing to do.
Anyway, it was a fun experiment. Whaddya'll think?