Magic is broken. Well, at least Commander is broken. Okay, maybe just for me. But I'll bet that some of you who've been playing this format for a while feel it too. I have a suggestion to keep you all occupied.
Shuffle your commander into your deck before you start the game.
That's it. Everything else stays the same. 40 life for group games and 25 life for duels. Use the Commander color identity rule for deck construction and the Commander ban list. If you buy a pre-constructed commander deck, great! It still works.
The real problem is variance. If you want to read an explanation of why, read on!
Variance vs. Consistency
I like my competitive games of MTG to have the least amount of variance possible. I want consistent performance from my deck. But! I want my casual games of MTG to have high variance. I don't want consistency. Too much consistency ruins replayability and the surprise interactions between cards. It makes the games too stale too fast. It makes it less of a puzzle and more of a procedure. You likely agree with this even if you haven't considered it. It's why the Cube format feels more dynamic.
Commander used to be a high variance format. A 100 card singleton format should be pretty darn interesting for a long, long time. But the entire idea of having access to your Commander over and over makes that not true in practice. And there are a lot of cards printed specifically for Commander now that are designed to reduce the variance.
Companions
For an example of this concept in practice ruining the game, look no further than the "companion" mechanic printed in the Ikoria (2020) set. This recent set included creatures that could be played from your sideboard if your deck met certain deck construction criteria. Sound familiar?
What happened in practice was that the companion mechanic broke formats from Standard to Vintage because it turns out having access to one more card than your opponent is too good. This is a resource management game. Starting the game with more resources than your opponent is always going to be good!
It turned out to be way better than they thought it would be.
Existing, cycle-tested decks started swapping a few cards to make a companion work. New decks were built around the companions. Basically, if you didn't have a companion deck, it was like you started down a card every single game. This exacerbated the problem with mulligans because non-companion decks would now be effectively two cards down off a single mulligan, further widening the gap before the game even started.
They "fixed" this problem by doing something they've almost never done in the history of MTG. They issued power level errata.
Power Level Errata
This tool is used very sparingly because it's confusing. In almost every case, if a card is broken, they ban it outright. Issuing power level errata means changing the way a card works so that it doesn't match how it is printed to "fix the power level." The rare time this happens is when a card is ambiguous in the rules or if the inherent structure of the rules change so that the existing card no longer works as intended. This happened with some of the very old cards as they've made changes to the rules over the years. The cleanest solution is to ban the offending cards. But that's not what they did with companions. This was an exciting new mechanic on a bunch of cards that had literally just come out and were still for sale.
Instead, they inserted a new rule. They said that instead of casting your companion from your sideboard for its mana cost as it was originally intended (and printed!), you could pay (3) mana to put the companion into your hand from your sideboard, then cast it as normal for its casting cost.
In other words, they added a speed bump to the cost of companions.
So on turn 3-4, instead of simply having access to one more card, you need to effectively "take a turn off" to move it from your sideboard to your hand. Typically this means that it will be in your hand for a turn, making it vulnerable to discard effects (for example), and it gives your opponent a turn to figure out what to do strategically because they know that it is coming.
Even with this power level errata, there are several companions that are still so good that people make the extra effort. It turns out that even paying (3) additional mana for a guaranteed card from your sideboard every game is a good enough deal that entire deck strategies are built around several of the most powerful companions. Which brings us back to Commander.
Commander
Having access to your commander every game creates a crappy sub-game where you're either trying to cast your commander as soon and as often as possible or stop your opponent from doing the same, or both. The rest of the deck in some ways doesn't matter. Those cards are just there to get your commander into play or stop your opponent from doing the same. The reason is that the commanders have become so powerful, they run away with the game.
An example is my Hope of Ghirapur deck. Hope of Ghirapur is a 1/1 flyer for 1 mana. In other words, because I can cast my commander on turn 1 every game, I will always have access to a flying creature on the first turn. It's relatively straightforward to add a ton of equipment to my commander and push damage through. If my commander dies, I can recast it for a total of 2+1=3 mana and go at it again with my equipment. This strategy is called "Voltron" for obvious reasons and it exists across multiple potential commanders.
This kind of tent-pole deck construction leads to very consistent games in a format that should otherwise be high variance. And the problem is significantly worse in 1vs1 duel commander where the entire focus is on one other player from the start.