Despite changing the sort order with the columns, when validation begins, it works from the bottom of the main GUI towards the Top.
The sort order has no influence on the order in which FV++ process entries.
I found out that FV++ is working in the order of the original MD5 file. The original MD5 file, a 13,000 item MD5 file that was originally created with MD5Checker, has the directories listed in reverse order (Z to A). When FV++ works with this file, the validation process also works in reverse order (bottom of the main GUI towards the top).
When I recreated a new 13,000 item MD5 file using FV++, the new MD5 file has the directories listed A to Z and in this case, the validation process works from the top of the main GUI towards the bottom (A to Z).