@Slaphead
can give you the best answer about how recording is done in practice, but I think that you are starting to visualise the concept.
Getting recording artists together to produce the original 'tracks' can be costly (imagine a recording combining 100 musicians, plus a large choir) so you would want to record as much data as you can on the day, and then you can edit the data 'at leisure' so to speak.
For a studio recording you could want one track per instrument, plus one per vocalist, and then you can add effects such as echo to each track individually, before you finally condense everything into one stereo mix. Also remember that for some albums, the instruments are recorded on one day, and the vocalists on other days, so in those cases you will want multiple tracks.
For orchestral recordings, the engineer may want to record enough tracks, so that the overall performance can be edited into various mixes, 5.1, 2.1, stereo etc - and this may require multiple microphones in several locations, with each recorded to a separate track.
So there are several reasons/needs why a recording engineer could want to use multiple tracks - Slaphead is the expert regarding the practicality of satisfying those needs.