There's a number of reasons to consider using arj as the archiver with tob. Although tob and afio are very well suited for each other, the biggest advantage of arj is that it will quickly and without fuss create size-limited, multi-volume archives.

To do that with afio, you'd have to do some chicanery with "split", which poses a number of problems for restoring and verifying said split archives. Lots of catting files together by hand.

arj will create multi-volume archives in the file system, and happily find and use the later files in the archive (.a01, .a02, ...) as long as they're in the same directory as the main archive (.arj). It also supports adding error-detection and -correction code to each part. Sounds great!

Under testing, arj was found to deal reasonably well when parts of the multi-volume archive are damaged. But if entire volumes are missing (e.g. you lost one of the CD's), arj will freak out and stop in its tracks, rather than skipping on to the next volume it can find. However, each volume in the multi-volume archive can stand alone, and arj will resume the restore if you execute a manual override and invoke arj yourself on the first of the remaining files.