> but you can't just claim "this effect must go away" without articulating a clear reason that it would.
If the effect is that, all else being equal, people with a lot of money can make more money in an absolute sense than people without a lot of money, then of course it doesn't go away. It's not that that isn't real (it's simply how percentages work) it's that it's not actually a problem because the model is so far removed from reality as to be irrelevant.
The real world is not a 1v1 adversarial game where people are betting against each other. More often they are collaboratively betting together and both benefit if they succeed.
Young, poor* entrepreneur brings an idea, maybe specialized domain knowledge, and time and energy
old, rich investor brings capital, maybe business experience and network, and gives it to entrepreneur to execute.
If all goes well old, rich investor and young, poor entrepreneur both make a lot of money. Young, poor entrepreneur becomes old, rich investor for the next generation.
If the venture fails, old, rich investor loses money (which went to pay some number of employees and vendors, who get to benefit from it), but old, rich investor expects this to happen for some or most investments. Young, poor entrepreneur loses time but gains experience and connections.
Nobody tricked anybody or stole anything from anyone or "lost a bet" like they are playing a rigged game in Vegas.
If you tell old, rich investor they aren't allowed to make any money by investing in young, poor entrepreneur any more, they don't just keep on doing it and allow you to redistribute their profits. They buy T-Bills instead. Young, poor entrepreneur goes to work for some other company (that old, rich investor probably funded in the past and owns) and gets a mediocre salary instead of getting rich and the world is deprived of whatever innovation they might have had.
This is pretty much fine for the old, rich investors, they're already rich. But it screws over the possibility of getting rich for anyone who isn't already. Which, if you were trying to reduce inequality, is the opposite of what you'd want.
* - or more realistically, middle or upper-middle class
Another unrealistic part about the model is that people keep betting a fixed percentage of their net worth. If you have a million dollars, maybe you can invest $100,000 into the seed round of a startup. If you have a billion dollars, it's unlikely you can invest $100 million into one investment. You spread it across multiple investments, maybe hundreds, and the average return is less than what you would get from succeeding on one big investment, because there simply isn't an opportunity that can make use of that much capital at once.