It kind of begs the question, though... other than maybe the initial capture while we play, why not use the FAR more sophisticated audio recording capabilities in the DAW in the first place?

Maybe you also read my thread on using the arranger's sequencer FIRST before recording to audio...? There are CONSIDERABLE advantages to capturing to MIDI first before recording to audio. One that I did not mention, but also bears fruit in a very noticeable way is the ability to transfer the SMF of your performance to the DAW, then record each TRACK separately (or as much breakout as you need), so that better EQ's, compression, effects etc., can be used as needed, and better audio automation can fix the little details that escape the onboard editing. Not to mention that you have access to MUCH better sounds, even, if you use sophisticated VSTi's to replace the weaker sounds in the arranger (this isn't a T4 slam, you can have weak sounds in ANY arranger).

If you are going to go the extra step and use a DAW, you might as well go the whole hog...