If every custom move were a side-grade, they would remain valid. They would remain valid because they change the way you play the character. By changing the way you play the character, it changes the way your opponent handles your differing application of the character.
This is what customs should do.
It is an unfortunate side-effect that some customs work better in tandem with eachother than others. This is unavoidable. But, if every custom had proper flaws to their strengths, the character in question using these moves that work in tandem, should have a much larger exploitable weakness. If the moves that work in tandem cover eachother's weaknesses, then this should hardly be a problem unless, again, the strengths are too great for them to be true 'side-grades'.
A weaker move that comes out faster. A stronger move that comes out slower. These are simple variations that many custom moves share. If they were properly tweaked and balanced so the speed and strength properly matched up and offered a different, unique, player-decisive change to the character, that would be infinitely more valuable than anything else the game could have to offer.
If anyone is arguing that customs are a way to make weaker characters strong, then it's unfortunate that these characters are so universally weak. I fail to see how pretending you can't change this, while simultaneously striving to prevent other people from changing this, is in any way a step in the right direction.
----
What if we applied a value to each individual custom move, based on how powerful it is, and disallowed "total scores" that are too high? (I feel like this was suggested before) Deliberately picking ****ty moves to offset the strong ones would prevent certain issues.
For example, if EBT/Trip Villager was forced to use the Lloyd that shoots straight upward, does that help address the issue when he can't properly wall you out after you break his camp? If not, what if he also or alternatively had to use the flower-planting non-pocket custom, which would prevent him from keeping projectiles out from his camp?
EBT/Trip doesn't look so menacing when his camp loses a dimension but retains its core gameplan.
This is an example, of course. We need more brainstormers to start looking at this, if only to pose the 'what ifs'.
It's actually impressively difficult to balance speed versus power in games like Smash, or even completely different games. The relationship isn't linear, and small tweaks affect a large number of variables. It's the exact same reason something that looks like a nerf to a move, especially if it's knockback or uncompensated damage, takes actual testing to judge accurately. If you make a move weaker from a damage or knockback stance, suddenly it has potential to become an unbearably powerful combo move, often with itself (looking at you, Sheik's fair or Mario's uair). The trend in a
lot of games, not just Smash, is that speed, versatility, and/or reliability will almost invariably trump power at higher skill. Things like Sheik's
relative lack of kill moves stop being relevant weaknesses when she has guaranteed or near-guaranteed setups, and top notch performance in all other aspects of the game. By contrast, no matter how much you buff Falcon or Warlock Punch in terms of raw power, they will never be considered overpowered. Even if they were ramped up to Brawl Minus Warlock Punch levels of power (i.e. 666% instant kill), the sheer unlikelihood of landing them on a skilled opponent would render them moot. Nobody gripes about Flare Blade killing almost everyone from almost anywhere from zero when charged, because there is no way Roy alone can force you to take that hit.
That's part of why we have so many weird power balances in Smash, customs or no. Setups and their likely or guaranteed followups outweigh power by a wide margin. Even with the Kong Kombo, DK hasn't gained the popularity of Zamus, Sheik, Mario, or even Ness, because his other drawbacks are so much more significant.
Many customs
do offer player choice. Most people grossly overstate the number of pure upgrades there are in the game. Even in cases like Pikachu, Heavy Skull Bash is only outstanding in how useless its base form is, like Flaming Lunge, or Falcon Dash Punch, or Warlock Blade. Obviously, some aren't in the running. Super Jump as a recovery option is simply too slow, still causes freefall, and has no invulnerability, to compare to a strong comboable kill move or a less strong comboable kill/recovery move. On the other hand, there are valid pros and cons to Fast Fireball versus default. Or about which PK Thunder to use. Or Crescent Slash vs Dolphin Slash. Things like Lightweight are "overpowered" because they give Palutena attributes almost universally considered vital to contend in Smash, while Counter is a situational joke in competitive play. People are afraid to give Sheik some situational Needle sidegrades when her potential benefit is vastly lower than allowing a Ganondorf to use Dark Fists.
It's nearly impossible for professional game designers to hit this balance. What individual or group within the Smash community would you be willing to trust with the task of trying to balance moves, or total loadouts? Who has the knowledge, experience,
and authority to define quantitative relative strengths to moves, and declare an un-exceedable cap? What skill level can be factored in? How much counter-play gets considered? Is EBT+Timber Counter (laughable now with the balloon explosion nerf) high-value because it's effective against characters or players poor at their own zoning, when a spike-happy Dorf player will gladly face such a technique? Is Kong Cyclone, the poster boy of jank, high-value against players who have the faintest clue how to play against it? Are we going to trust a bunch of biased elitists (speaking of both sides, self included) to rank these objectively?
I don't think anyone in the community has the right to make such a ruling. And it would immediately lead to dissension.