> How exactly is the preference of the one individual dominant?
The other 999 people will have a diverse set of preferences for services and goods within the economy. They will share some, and not others. But each of them has only $500 to spend on those services and goods, and so anything not widely shared among them doesn't offer much of an incentive in terms of potential income to anyone who might offer goods and services. If something doesn't appeal to a sizable majority of the 999, there's likely to be too little chance at sufficient revenue to pursue it.
By contrast, the high value individual has $500,000 available to spend on their own singular preferences for goods and services, providing a strong incentive for providers of those things to cater to this individual. The changes of being to have a reliable, significant revenue flow from doing what they want are quite good.
Consequently, the economy ends up being structured (to some degree) around the preferences of the wealthy individual, rather than the broader public.
Of course, this is not a "hard divide" ... obviously there will be some things that all 999 others (and even the 1 rich individual) want, and some part of the economy will respond to that. However, even for such basic things as, say, bread, preferences can be wildly divergent, meaning that a single provider is not likely to tap into the entire potential market for that generic. And obviously there are also risks in this thought experiment scenario, because there is only a single high wealth individual - fail to accurately target their preferences, and you're potentially worse off than catering to everyone else.
Obviously, a more realistic example would (e.g.) involve 1% (say, 3M) individuals of high net worth/income, amidst 330M+ individuals of varying levels of net worth/income. In this scenario, there is more scope for providers of goods and services to cater to relatively unpopular things, because amidst the 330M lower-income folk, there will likely be enough to make it potentially worthwhile.
But this is true only to a point. If enough of the lower income folk don't really have much disposable income at all, their preferences for non-essential items will not translate into actual purchases of goods and services that they may otherwise be desirous of.
And of course, with 1M high income people, their preferences will be more diverse than in the thought experiment. However, the deck is still stacked in favor of businesses that seek to cater to those with more money, because of the unbalanced distribution of wealth & income.
In this more realistic example (i.e. the contemporary USA), a sizable chunk of the economy remains focused on the preferences/needs of the majority. But a much bigger slice of GDP flows around the preferences and needs of the very wealthy than would be the case with a more balanced distribution of income/wealth, and I maintain that this constitutes a harm to the economy and the entire population.