No one can tell you what quality one setting is over the other. However, we can tell you what the general census is. -V 0 is generally regarded to have the same quality as 320kbps except in rare cases or with killer samples (every lossy encoder has rare cases and killer samples though). It is said that -V 0 is better because it can achieve high quality with the files being much lower in size than 320kbps.
That being said, -V 2 is regarded as being the standard in that most people (with most samples) can't distinguish that between the lossless source. If you want my personal opinion then it is this: I would only download at -V 2 given that it is definitely transparent to my ears and the file sizes are even smaller than -V 0.
I suggest that you conduct a blind ABX test with Lame mp3 at various bitrates. Compare different Lame settings (start with -V 5 and go all the way up to -V 0) to a lossless file ripped from a CD. You probably won't be able to properly distinguish between -V 3/2 and the lossless file.
Your last statement about the base, hi-hat (a high hat is a hat high in the air

), and snare need to be backed up with ABX tests. Otherwise you are probably just suffering from the placebo affect. The placebo affect occurs whenever you know a bitrate of a song and you think that you are hearing things that you aren't actually hearing. For example, you view 192kbps as being a low bitrate (it actually isn't a low bitrate, that is a high bitrate). So you listen to 192kbps music and you know that the bitrate is low (again, from your point of view). So then you start to hear all these flaws that aren't actually there just because you know the bitrate is "low."
Remember that bitrate does not necessarily reflect sound quality. It is true that a 320kbps CBR Lame mp3 should technically have the highest sound quality that Lame has to offer in the normal mp3 bitrate range (ie not free format). However, there is always a point for nearly everyone out there in which bitrate stops making a big difference. Many people cannot distinguish -V 2 (~190kbps VBR) from the source lossless file so anything above that would be a waste. So in this case, -V 2 would actually be superior to 320kbps CBR as -V 2 not only achieves the same quality but has a much, much smaller file size.
I hope all that helps. You need to conduct a blind ABX test before you can make any serious comments about audio quality though. Otherwise your comments won't be taken seriously and you are violating the TOS here at Hydrogenaudio. Remember that the point of Hydrogenaudio is to correctly guide people. If everyone came on here saying that "the bass can sound bad" without actually having proof then we would all be lead to believe that Apple lossless is not lossless (it is in fact lossless, I have just seen quite a few posts where people say that they can hear a difference without ABX testing), lossy encoders (Lame mp3, Nero AAC, iTunes AAC, etc.) suck, and people should use old fashioned tube amps and vinyl for everything.