@ctheory11 Well, to answer to your question 1 : it's just to give more work to mastering engineers since they will have more tracks to work on !! Hey, budy ! Just kidding here of course...
My point of view to answer both questions is : to get more freedom and sound adjustment levels.
1- Though you can put all your individual drum's sounds in one Drum Rack and all the Fx chains on each cells of this Drum Rack to sculpt each sound individually, you quickly end up with a not so smooth midi track to deal with when you want to make changes on the fly. That's already a visual & workflow issue. Next to that, unless you route the audio out from each cell from your Drum Rack to individual audio tracks, you're ending up with all your drums coming through only one stereo channel which is not giving you the same level of controls for levels, panning, phase...etc. Think real life situation where a sound engineer will record a drum set in a studio : he's going to use multi-mics to record to have some freedom during the mixing process afterwards. Then the fact to group all those tracks give you the common sub-group / Bus option to control the level and fine tune the all drum kit before reaching the master or the mix-bus. So I think that's good practice and that's without talking about Live Audio & Midi workflow, the first & last fader you want to adjust is your track level's fader, just for gain-staging. Any device, plug-in that you put on a track is acting on the sound before that one. So you sculpt the sound as much individually as you can before reaching that stage and this method is giving you plenty of freedom in that regard.
2- Same principle, more freedom and flexibility for sound sculpting. Using controls like attack, decay,sustain, release, volume...etc gives you much more flexibility on your sample playback and also more interactivity with your Midi notes events like length & velocity. You have more restrictions / Less options when using the One-Shot mode.