That actually follows from the traditional argument against possibility of welfare - if the state can do such help, it'll first give it to closest to it, which are the people who need it the least.

But I think with direct democracy it'd be fine. At least some middle ground would be found between those voting for "free money" and those voting so that others wouldn't get "free money". Unlike now when depending on who you are it's either always free money or always fuck you.

EDIT: In general radical political models are better thought through fundamentally. Real world ones work in arcane ways, usually not the ones publicly declared, and rely on lots of inertia to be functional. But both radical marxism (direct democracy and full on social involvement) and radical ancap (no common decisions at all, no common social involvement at all) lack such vulnerabilities. That's unfortunately the reason people with real world power don't need them. If you have real world power, you'd support the change that gives you more power or preserves what you have. So for a model to be plausible it needs to have vulnerabilities, to attract real-world support. Only disadvantaged people really want a perfect model, and they are not the ones deciding.

Hence another radical variant - radical agnosticism of political systems, try to always keep as variable and diverse mix as possible, so that power, advantage and disadvantage were more or less equally spread, allowing people to live maybe not in heaven, but not in hell too. Decision-making systems as mixed as possible, legal spaces as diverse as possible, and so on.