I don't buy the argument that government intervention is the reason for low wages. The implementation of the minimum wage in the United States was directly related to the fact that capitalism without government intervention is serfdom. Back when the "free market" was allowed to set the cost of labor, people worked 70 hour weeks in dangerous conditions and still lived in extreme poverty. Industrialized countries without government intervention in the labor market have sweatshop economies, without exception. The real problem is that corporations own government to a substantial degree (at least in the United States), and so the minimum wage has not kept pace with productivity or with the cost of living. To be clear, I'm not saying that going to a UBI as opposed to an increased minimum wage would produce a sweatshop economy. But I am saying that low wages are the result of insufficient government intervention.
I'm not particularly worried about a minimum wage increase incentivizing automation. I hope that it would incentivize automation. I don't see any particular value in having a human waste 8 hours a day doing mindless repeatable tasks that a robot could do just as well. When automation gets to the point where there is no longer enough work for people to do, at that point UBI will become necessary. Given that there's currently more than enough work to go around (especially if we add in all of the jobs that could be created by a Green New Deal type of program), it seems prudent to further incentivize work rather than make it easier to get by with little-to-no employment income.
Andrew Yang's UBI proposal does effectively eliminate government programs. Few, if any, people receive over 1000 dollars in government assistance from programs that don't stack with UBI. Which sounds good at first glance - that means low income people get more assistance than they're currently getting. But let's say you have two people. One makes 20,000 a year, has 300 dollars in government assistance replaced by UBI, and pays an additional 100 dollars a month in VAT. The other makes 80,000 a year, has zero dollars in government assistance replaced by UBI, and pays an additional 200 dollars a month in VAT. In that example, UBI has a 600 dollar a month benefit to the 20,000 dollar earner, and an 800 dollar a month benefit to the 80,000 dollar earner. In order for wealth to be effectively redistributed by a UBI program, it needs to stack with all government assistance and/or it needs to be funded in a way that affects the wealthy proportionally more than it affects the poor. In fairness, it is possible that the second point could have been true under the Yang plan, depending on the VAT rates on staple goods vs. luxury goods. That would complicate the MATH, potentially in a good way, but I don't think the proposal ever got that specific.
I don't really see how a wealth tax punishes people for success any more than any progressive tax. As you say, people who spend 120,000 dollars on VAT eligible goods are fine. It strikes me as inconsistent, given your apparent understanding of VAT as a tax that would disproportionately affect the rich, that you don't consider that a punishment of success. Beyond the inconsistency, it's a matter of objective fact that wealth distribution in the United States is insanely inequitable. Wealth must be redistributed downward, which means that, for the benefit of everybody else, the extremely wealthy have to get used to sitting on slightly less treasure.