These are the fallacies that I'm highlighting.
1) Unions will not permit the dissolution of the systems in place.
2) 'Means testing' should not theoretically be hard - and - it's actually a rational thing to do (so long as you point out the disincentive is not there) - ergo a reform of these programs should yield much the same benefit. But they won't and they don't.
So the first fallacy is to not recognize that the means by which these things are organized has less to do with economic theory and more to do with organizational competence and efficiency - which we struggle with.
But on the other side of the equation:
3) We don't understand the nature of work, poverty etc.. Everyone thinks about homeless people as though they have the same life motivations as the rest of us, grinding away in the system, and that's wrong.
People in tent cities are already mostly opting between 'tent city squalor' and 'Amazon factory drudgery'. Yes, it's hard to get a job from a tent, but even they had the chance to work at the Amazon / Waffle House, they just won't want to do it.
Those people will not come out of tents for $500 a month even with the opportunity to earn more. In an absolute free market situation, they would live in tiny, semi-permanent structures and chill out.
When you mix drugs into the equation, incentives are more perverse. Where I live we have $1K a month means tested welfare, and many homeless opt for living rough during the summer (and taking a flat in the winter) because they'd rather spend it on crack. That dynamic is only exacerbated with more money.
Students fresh out of College, in the area that I live in all have a lot of side interests. I know many of them and I believe the average choice among them would be to take UBI, work on their documentary (which is not very good), go to the gym, and maybe work a few days now and again (serving coffee) for some money for the trip to Africa.
With free Uni, free Healthcare, almost free childcare (this is Canada), you've subsidized the entire cost of living and externalized all of the difficult work, ugly work, and especially the kind of work that takes a lot of time and professionalism to develop. Esp. those 'late stage career people' with significant responsibilities upon which we depend a lot aka 'the legitimate elite'.
Which is where the reality of the economics comes into play ... even if someone could make the argument that UBI 'is affordable' - which I would totally disagree with other than for a paltry amount, say $500 a month, the loss in productivity at the lower end would be substantial, to the point wherein we'd see large swaths of value evaporated, the cost of labour would rise significantly and incentive structures would start failing.
If we're talking just 'economic incentives' than probably just better taxation at the higher (getting rid of loopholes and higher rates on the ultra-rich), and a bump in mimimum wage would make a big difference because the allocation of that 'extra capital focused on the lower end' would still have some market efficiency and it would be oriented towards productivity.
That - and basic, communitarian social measures would work.
I believe UBI is like Crtypo or Bitcon - utopian solution that fails for fairly obvious reasons, like a kind of 'solve the equation' thinking that we technical people have that just doesn't work once you account for the more obvious social features of society.
I don't think there are any magic solutions. We have to raise our kids well, provide some kind of quasi-level playing ground, and accept some people will fall of the edge and have some way to help them and to try to minimize that.