I like that this theory/rule spreads around the byes so that the same player is not always affected, and the logic for selecting the player is solid.
In my tournaments I can see this affecting prizes. My three-round swisses can have nine players. If player #5 is given a bye and the stronger players win in round 1, then #5 could be paired down in round two, and two points is usually enough for a prize. It is also a realistic possibility for player #5 to be paired down again in round 3 if he loses round 2. The road to two points is more difficult for his peers with similar ratings.
Somehow, I have the feeling that this entire conversation would never have become so long or detailed if more TDs would try just a little harder to avoid byes in the first place.
I hate byes. When I’m TD’ing I’ll do everything reasonable to avoid them – cross-round pairings, cross-section pairings, or the TD will jump in and play whenever necessary.
Concerning the latter: When the TD jumps in, he should not pair himself against the player who would have received the bye. Instead, he should pair himself normally, according to his score and rating. He should assign himself a “pairing score” for his unplayed games from previous rounds, i.e. he should put himself into whatever score group he would be about in the middle of, rating-wise.
This is what I try to do, time permitting. If I have a 7 player 3 round Swiss, I give the bye to the lowest rated in round 1 but then pair him/her against the first loser. The first loser is then not paired in the next round as this new game becomes a cross round pairing. By doing this, I have only one bye, instead of 3.
As this would be done in the lowest quad only, there’s a good chance that those games will finish quickly so that the section can still finish on time as lower rated players usually play more quickly.