Correct, but what I am describing is a wealth tax system that goes after unrealized capital gains from stocks that does not depend on the stock being sold. It can tax the expected value increase rather than the actually value derived when a stock is sold, demonstrating a working method to tax the value of stock ownership rather than at the point of sales. It also demonstrate that people will volunteer to such system if it gives a statistically economic benefit to do so.

If they are taxing unrealized gains in one year, are they returning funds at the same rate on unrealized losses in the next year? (This isn't a rhetorical question. I'm genuinely curious.) I don't understand how this is going to work.

No, a predetermined rate is predetermined regardless if the outcome at the end is a loss or profit.

The benefit of doing so for the investor is that the potential for profits is higher, and on average the predetermined tax rate is a bit lower than the actually tax you would pay after selling the stocks.

Let say as an example you bought stocks in an index fond between 2010 and 2018. The average result from all index fonds was +8%. With regular income tax on sales the resulting profits people got after tax was 5.6% each year. With the predetermined tax they got to keep 7.547% out of those 8%.

But this year the crash might have resulted in less than 0.457% profits, in which case the predetermined tax can be higher than what the income tax would be on sales.