You shall not re-encode from mp3 to mp3.
Joint stereo is theoretically better, even when transcoding from forced stereo, but at those bitrates (-V2) the difference is unlikely to be audible.EDIT: greynol beat me.
Post-processing such as Dolby surround decoding or karaoke-style applications.Other than that, we've already answered your question with all the information you could possibly need.In case it helps, let me put it even more plainly: use joint-stereo even if the original file was in stereo.EDIT: pdq beat me this time around, lol.
Your question has been previously answered regarding the use of stereo and joint stereo (hint: its the second option). You were pointed to the Wiki articles as they contained information to help you determine the proper settings along with gaining insight into what you were asking. People often come on and ask "What's the best setting for X encoder and A, B, and C type of music?" There isn't an absolute answer for anyone. However, it is generally best to use the default Lame settings unless you are experiencing problems (that can be heard in proper blind ABX testing) and insist on tuning them out. Lame is also a great VBR mp3 encoder and you can likely get away with not using such an insanely high bitrate. Again, unless you can properly ABX it, 320kbps VBR is probably going to be overkill. I suggest starting out with your source lossless material and use the -V 5 setting (maybe even -V 6). Conduct a blind ABX test with a few songs comparing the source lossless versions to their lossy counterparts. "Pass" the test and encode the material at a higher setting. Keep progression (-V 6 to -V 5, -V 5 to -V 4, -V 4 to -V 3, etc.) until you fail the test. Here is is a link covering the ABX process:http://wiki.hydrogenaudio.org/index.php?title=ABXThen again, I know how you hate getting links as answers so I am not expecting you to read it.