I believe destroy buyout to be a good example of the inherantly flawed idea of preventing natural steamrolling in this game as it removes, from the players perspective, his agency and ability to make decisions which allow him to win.
Agency is an interesting word. In video games it's often used to talk about a lack of agency, a lack of meaningful options available to the player, many times used to describe a story in which no matter what you do you follow a linear path and change nothing, you personally had no effect on the story. Ofcourse, in reality in single-player games the players really do have no agency, everything they do must have the game be able to respond appropriatly , otherwise the game itself breaks down. These games instead make the player feel he has agency, by making the game so good you don't want to actually deviate from the linear path or by guiving the player so many chocies that he feels he's changing things or even by simply keepign the pace of the game moving fat enoguh that they never stop to question if there could be a different choice to even take. However, this is all single-player. In competitive multiplayer you actually do have agency in the real sense of the word, your choices matter, what you do has a direct effect on the outcome of the game, to win or to lose. These choices are countered by the opposing playes, and the person who makes the better choices wins. This is what competitive games ultimately boil down to, and try to achieve the best they can whilst also being fun both for those who lose more than they win and those who win all the time, to keep them playing. This is often why random elements are added, to 'spice things up' or make a fight more even or sometimes just to give the losing players readily available excuses. I have no problem with this.
What I do have a problem with however, is when you have no real control over whether you win or lose, again not talking about RNG. In destroy buyout, when you buy someone out their stuff is wiped from existence and you gain claims depending on what their HQ level was. This is designed to make snowballing harder because one criticism of this game has always been that if one person gets a quick buy it's basically game over for everyone involved. What I feel this mode does however is go the complete opposite way, a buyout is not worth enough for it to be worthwhile no matter what stage of the game you get a buyout at. This would be potentially an easy fix, give the buyer a higher stock price, give them more claims etc. but I am beginning to think that doing this doesn't address the main fundamental flaw of this mode, not steamrolling causes situations in which you lose agency.
Imagine a situation where player 1 has bought out player 4 and is now the biggest, with players 2 and 3 being equal. In order to win the game players 2 and 3 likely need to beat player 1 first, before he can grow enough to just buy them both, however if player 2 grabs player 1 then player 3 is left with a lot of cash, whilst player 2 only has a bunch of claims. Player 3 wins because he has such a cash advantage that the stock price increase of player 2 was not enough to win. Claims are interesting in that they provide resources at the cost of tempo, it takes time and money to convert those claims into things producing money. Because of this, player 2 will never be able to get the new claims working quick enough to overcome the large cahs stockpile of player 3 and win, and there is no surefire way to attack a players cash, and there shouldn't be. So, smart players will know this, and instead players 2 and 3 will attack eachother, but wait, that means player 1 survives long enough to buyout player 2 or 3 when one of them takes over the other. So for both players 2 and 3, there is no winning, player 2 has to make the choice, does he want player 3 to win or player 1 to win? Neither option is a good outcome for him, so he cannot make a decision. He has no agency in this game anymore, his victory depends entirely on other people's decisions, for player 3 wins if player 2 decides to eat player 1, and player 2 wins if player 3 decides to do the same.
Now let's look at it from player 1's perspective. You see, whilst not as bad as the others, he also has this same problem. If he attacks player 2, player 3 will be strong enough to eat him, and if he attacks player 3, player 2 eats him. His only option was to have had enough money to be able to buyout both player 4 and another player at the same time, before anyone else can get a cash stockpile. That very rarely happens in destroy buyout. So player 1 also has no agency, his actions only let other players win. The situation is a deadlock, whoever strikes first loses, so no one can strike. The only option is to get enough cash to buyout the other 2 simultaneously, and only player 1 can do that since he's the biggest. So, if no one trikes, player 1 wins. Players 2 and 3 both know this, so they must both make the decision, 'do we let him win?' Player 1 is likely more skilled than the other 2, he got to a buy before they did. In an ideal world, as soon as player 1 gets a buy both of the others know the game is lost for them, so player 1 wins by default. Is this preventing steamrolling? But wait, player 1 can lose, so it prevents steamrolling... except then your not rewarding skill. Is it truely fair to have the winner be decided by the losers choosing who they want to lose to? No, no it is not.
The reason for this issue is simple, claims have no real instant value. A buyout rewards you with nothing. Without destroy buyout, you buy someone and you get all their buildings, their infastructure. This has real value. 4 steel mills is like getting 40 iron and 20 seconds worth of tempo, the build time of those steel mills. This is what leads to snowballing, this effective multiplied many times over. Without destroy buyout, you can grow quickly enough that the game is basically over as you assimilate everything that person had into yourself. With destroy buyout however, your not gaining tempo for the buyout, your losing tempo because you have to put those claims to use. A well known formula that could summarise this is time = money. Buying someone takes money, which is time. When you get their stuff you gain a lot of the time back. Without this stuff, a buying player just loses so much time that they cannot possibly win, unless their opponents agree to let them. Because of this, the idea of destroy buyout is inherantly flawed. If you give them more compensations, they will just steamroll again, if you don't, well in the ideal world that didn't matter, they still won as soon as they got that first buy, but in reality the buyer has to have his opponents agree to lose. If you don't reward buys at all... well what's the point of buying then?
TL;DR: Getting claims for buying someone prevents steamrolling the loosest sense. A steamroll doesn't happen only because the losers can choose who they lose to, which is not rewarding player skill, so in an ideal world they will choose to lose to the person who makes the first buy, which means they just steamrolled afterall.