I feel like this could potentially work with "voting tokens" or something else that's not money. Maybe with some sort of crypto verification. You'd still have the scarcity that the system demands, but without the "inequality baggage" you'd start out with if the system were implemented with actual money.
If you implemented this system using money to buy votes, then you'd end up... well probably with something like the inequality situation we have today.
I think one feature of the system is that the accumulated funds are then distributed equally after the vote. With tokens, this would not be very useful, as if everybody gets the same amount tokens, there's no difference between getting one token and getting one million, tokens themselves have no value.
If voting tokens do not perish between votes, that leads to the similar inequalities if I abstain from voting for a number of cycles. If old tokens do expire, then it doesn't matter when everybody else gets the same - up to the limit of the quantization of tokens, how many tokens you have is less important than how much more tokens than your opponents you have . I.e. if there is a yes/no question, the important thing for you, if you favor "yes", it to have more tokens for "yes" than for "no", but how many tokens for each does not matter. In multilateral choice it's more complex but I think if you have enough tokens to not be limited to voting for one thing because you only have one token, it is again the same situation.
The whole point of the system is to railroad people into only voting for what really matters to them, by "saving up" their votes. I think this isn't a flaw but a feature.
I'd expect that there'd be a fairly large number of tokens so that quantization wouldn't come into play too much.
If you implemented this system using money to buy votes, then you'd end up... well probably with something like the inequality situation we have today.