1) I'm not using groups (other than AP and Controllers), so I can't comment on this.
2) Not sure. You can schedule the jobs in Airwave, so not sure if scripting would provide anything different.
3) My experience is yes. There is an option in the Firmware upgrade job to allow downgrade (sorry, not sure of the exact wording), so I believe that pushes the firmware to the unused partition....or partition with the oldest firmware (or maybe the current non-default boot partition).
4) I'm on my 2nd/3rd upgrade cycle using Airwave (from 5.0.2.0 to 5.0.3.3 to 6.1.3.x) and have not had any issues with bricked controllers.
My only issue has been jobs that get stuck. Seems like Airwave will only initiate 1 transfer at a time. There is a setting to fail after X amount of times. I have mine set to 1 (believe that is default), but I have noticed that I'll get an occasional controller that will fail, and the job will attempt multiple times (seems like once an hour) before finally failing and moving on down the list.
First time that happened I had controllers upgrading in the middle of the day (my fault since I forgot that I had kicked off the jobs the night before). Now I set the job and wake up early to check the status. If anything gets stuck I just cancel the entire job and try again the next night.
Haven't figured out why some controllers get stuck. Error log usually indicates bad password, or file transfer failure, but all my stuff uses same credentials, and subsequent efforts seem to go through, so may have something to do with my network connection to those offices. Stubborn controllers get upgraded manually via FTP......but really, the problems have been very few and far between.
Hope that helps.