Like I said, you can make a historical argument. Though tariffs were usually based on the value of the goods, so my argument applies there, and land sales (as far as I'm aware) were done by auction, so do not represent net demand for dollars. Not to mention that for most of the 19th century we had some form of hybrid metal backing for the currency.
But in modern times, ~90% of government revenue [1] is from taxes that are relative to income, and thus create no net demand for USD. I don't know when this became the case (at least the 60s), but I suspect for a very long time. The federal government, as far as I'm aware, has no property taxes; those are at the local or municipal level. And there, they are based on the assessed value of the property, so if dollars are expensive, then the taxes are lower, and if dollars are cheap, then the taxes are higher. Because the liquidity is required, you could argue that there's some marginal effect, but given that equity is easily converted to debt at scale, I'm dubious that that would ever be a factor.
In my opinion, fiat money is the closest that we've been able to come to "good" money in a sense -- easily transactable, scarce, fungible, hard to counterfeit, easy to validate. The value derives from these characteristics. The "scarce" aspect is the one that bothers me the most, because it doesn't have the same guarantees about scarcity as something like gold (although it excels at the other points). The notion that money is "backed" by anything is and always has been a red herring in my mind; in almost every case, disconnects between the thing that explicitly "backed" a currency and the currency itself have had little to no medium-term effect on the value of the currency, because the currency typically "inherits" the scarcity aspects of its backing. To what extent cryptocurrencies can succeed is still up in the air, but I remain very optimistic.
[1] https://www.cbo.gov/about/products/budget-economic-data