I had always just assumed it must have been an ease of implementation thing. (Explanation 8.) Their suggestion that you could just check the 10s column instead of the 1s is misleading. You would have to test both the 10s and the 1s column to get a 10 minute differential, unless the original alarm time happened to be a multiple of 10.