For organizing my music library I use a modified version of the application foobar2000, as well as a bunch of other third party software for editing tags, converting audio formats, etc.
The folder structure goes somewhat like this [Genre] / [Album Artist] / [Year] - [Album Name] / [##] - [Track Name]
On Discogs and last.fm I look up a lot of the relevant information for the album, like release year, country, record label, album artwork, style tags, background info (so that it can all be searchable and easily identifiable). I follow certain structures, like if an album is a compilation album released by a label and no particular artist, I make the album artist the name of that label and group them together that way. For the genres and styles I use the modern id3 v2.4 standard, which is not widely supported but allows for multi-genre and multi-artist tagging. The first main categorical genre determines the genre folder it lands in. All the other genres and styles are stored as extra descriptors so that everything can be browsed and queried using foobar's wonderful adaptive and re-programmable user interface.
Using id3 v2.4 tags, I neatly write all this information to the files and adhere to a standard for the whole library and make sure that I have at least between 600x600 and 800x800 pixel size artwork
My modified version of foobar displays a waveform seekbar, allowing me to visualize the current listening position of the song super imposed on top of the peak waveform of the whole song. This gives me extra insight when studying the dynamics of music and instantly allows me to see how a recording was mastered. Also, there is a high quality VU (peak and rms) combined meter, loudness normalized around -18LUFS using the EBU-R128 loudness normalization algorithm. This meter gives me a relativistic perspective on peak and rms dynamics and mastering quality of a song. Aside from this I have a very large display of the album artwork for more visual contemplation as well as a real time stereo fourier transformation of the audio signal, allowing visual cortex analysis of the temporal stereo frequency spectrum changes in the music I am listening to.
Of course, I only listen to everything using ASIO soundcard drivers. The windows sound module alters and muddies up the sound as it is known to use weak resampling algorithms to mix multiple sources of audio together, myself and others have confirmed this through listening tests. When I listen through direct ASIO driver output I get more clear sonic quality and I can properly make out all of the finer details in recordings. Often times, I think most people don't realize that the difference between playing back a 24 bit lossles audio file and a decent quality mp3 file is negligible compared to the difference in sonic quality you get from switching to proper ASIO soundcard drivers. In other words, if you're not even using ASIO drivers, who cares if it's mp3 or wav? I say, make sure your digital signal processing chain and audio equipment is set up proper, then worry about which compression algorithms to use. In my experience, I am fairly happy with high quality mp3s for my music library; but of course I only work with at least 32 bit floating point wave files for all my serious audio editing work and do my calculations in 64 bit with 512x oversampling.
Using my knowledge of computers, mathematics, and programming, I have learned certain truths about computer digital audio over the years and combined all this knowledge into my customized workflow with audio applications to really allow me to study and appreciate music in all its details.
That's it for my music library. I have separate libraries for percussion hits, audio samples, recordings form portable recorder, etc. I primarily now use Reaper to compose, arrange, edit, mix, and master all my audio.
Reaper is like a swiss army knife for the studio work and foobar2000 makes an excellent music exploration and studying tool ~