No. Assuming you get to CP for set point, you will always have the option to choose the stage you first won on (unless your opponent takes you to your first won stage and loses there, giving you 2 wins on the same stage anyways).
I think the underlying problem is that only one player will be able to CP for set point, and that player will always have the choice to go back to the first stage they won on. This is an inherent issue in best of x sets, and could only be solved by making sets be won by 2 games or more.
The only way you can cp the first stage you won on is if you won game 1 or if you won on their cp. Example of sets when Falco can cp the first stage he won on vs. Marth (wins on Marth's cps are italicized):
Falco wins on BF. Loses on FD. Wins on DL. Loses on non-FD. Can go to BF again for game 5, but that's not an advantage since it's the most even stage (according to the players).
OR
Falco loses on BF. Wins on DL.
Wins on FD. Loses on FD. Can go to DL again for game 5
OR
Falco loses on BF. Loses on DL. Wins on DL.
Wins on FD. Can go to DL again for game 5.
If I'm missing a way for Falco to play DL twice without winning on Marth's cp, lmk. Either way, as you can see by the stage choices, DSRm can frequently result in 3-stage sets as opposed to BSR which will typically result in 5 stages being used. With DSRm, players will cp the best stage they can, which is completely dependent on the order in which matches are won (arbitrary, and should have no bearing on what stages get played). With BSR, players will cp their best stage after their first lost, and their second-best stage after their second lost, or vice versa. Obviously it doesn't always happen that way because sometimes they may want to go back to the stage they struck to instead, but the stage they struck to is perceived as even, not as an advantageous cp. See the last quote.
When you make a ruleset you have to think about its actual usage. Rules that require interaction from a party other than the 2 players (usually a TO) are inherently disruptive to the tournament. Rules that have overhead associated with them add additional work for the TO staff, which is inherently disruptive. TOs have more important things to do than baby sit the players. When these rules are scaled to larger tournaments, it takes exponentially more work. For that reason alone, the rules will have a hard time gaining popularity.
Determining an infinite loop is difficult, sometimes even for computers. Operating systems usually use a timeout or counter to guess whether or not there is an infinite loop (hence why windows prompts you whether you want to end a program or wait for the loop to finish... it doesn't really know, it just knows its "been a while"). I would suggest changing your rule to something along those lines. If the players can't agree within 5 choices, they will play their default characters. For practicality, it might be better to simply allow the players to choose their first character each set, with or without a blind pick.
And as a TO i have been called over for double blind picks on occasion. It does happen and it is important.
You say rules that require a TO are a hassle, which I agree with, but you go on to state that you have been called to settle double blinds as a TO. So how is my method any more of a burden? You're just assuming that my character selection method will result in a bunch of infinite loops even though this is already the way most character selections go. Ex. Mew2King sits down to play someone and selects Marth. The opponent goes Falcon, so he switches to Sheik. I've never seen someone request a double blind BEFORE they know what their opponent's character choice is. People call a double blind when there is an infinite loop, but they can also do it to bluff people out of cping them. All my method does is protect people who use multiple characters from literally having to guess on the double blind.
Determining an infinite loop isn't hard. It happens when the players come full circle in their character decisions. Ex: (Player 1 vs. Player 2)
Fox vs. Marth
Sheik vs. Marth
Sheik vs. Puff
Fox vs. Puff
Fox vs. Marth <-- Obviously they will just continue cping each other unless one of them changes their mind about what matchup they are willing to play. If P2 for some random reason decides going through the second iteration that will be actually okay with Fox vs. Marth or Sheik vs. Marth or w/e even though he wasn't originally, then he can just continue the loop and stop it on that matchup. The best part is that the players can just go off of each other's word for what their default is, and you can just check when you report the score to the TO. "I lost 2-1. Was ___'s default character Fox? Oh, okay, just checking." And if they find out the person lied about their default character, obviously they will just be DQed.
Fox vs Marth
Strike to YS, marth wins
...
Fox bans YS, marth picks FD
...
Fox bans FD, marth picks YS.
So your rule doesn't completely solve the problem. Also, it eliminates true stage bans while creating even more overhead for the players to keep track of. Anyways, as you said, the point is to limit hard CPs, but regular stage bans is the process of removing hard CPs altogether. You can ban FD and not go there even once. Without those outlying stages, the players will be forced to choose from the more "even" stages and thus have a better set.
The problem with this is that if you strike to YS, you can't possibly complain about playing it again... It is very frequently the case that you will have to play on the stage you struck to for rubber match of a set. If a player somehow strikes to 1 of their opponent's 2 best cps, there's not much the system can do to help them at that point. The Fox obviously considers YS an even stage for the matchup, so the fact that Marth is better at the stage seems irrelevant. There's no abuse of cps occurring. It's just Marth and Fox having differing views of what stages are good for the matchup, and Marth's understanding of the stages/performance on the stages perceived to be the most even is just greater than the Fox player's.
