IMO, add-ons are for cash games, not tournaments.
Re-buys are great for home games because they open the door for players to come back into the game for a defined period of time and they inflate the prize pool. Top heavy structures that pay out a generous portion of the field incentivizes players to buy back into the game when felted. As a case in point, the house would keep $75 in the $50 thirteen to fifteen player tournament I use to play in. It was not uncommon to have a couple of players in for three bullets - $150. The pay-out structure was 45/30/15/10.
A third of the money would come from re-buys. Twenty-one buy-ins would generate a prize pool of $975 after the house took out it's share. For simplicity purposes, the host would round up or down so the pay-outs were divisible by $25. Typically, players finishing on the bottom half of the pay scale were the ones shorted by this formula. A tournament with 21 total buy-ins would pay out: $450 for first, $300 for second, $150 for third and $75 for fourth. Which begs the question, why not buy back in when you can potentially win $450 or $300?
That brings us to the one consideration casual players tend to overlook to the benefit of stronger players when it comes to re-buys in home games and that is the law of diminishing returns. A player who is in for $150 and is lucky enough to finish in first will win three times his $150 investment and twice his investment for second. He breaks even at third and recoups half with a fourth place finish. Conversely, a player who finishes in the money with his original buy-in respectively wins 9x, 6x, 3x and 1.5x his money. Eliminate re-buys from the equation and that same player who won $400 on his $50 investment for his first place finish now nets $250. (Effectively, second place money.)
You asked.