> UBI doesn't really make sense as a replacement to almost any traditional welfare programs for the simple reason that by limiting the money to only those who qualify for it, you can give more.

For means-tested programs, you can replace means-testing with unconditional benefits and appropriate progressive tax rate changes, and acheive exactly the same effects (modulo differences in definitions of income, but is there really a big benefit there?) while eliminating duplicate bureaucracy. Most UBI proponents don't prefer as rapid of a claw-back (which sometimes exceeds 100% in some ranges) as the aggregate of existing means-tested welfare programs, and would instead claw-back at a slower rate but extend the claw-back higher up the income distribution so that the drag is on the rich rather than the lower middle class, to improve upward mobility compared to existing means-tested welfare, but if you wanted to optimize only for bureaucratic efficiency without moving the claw-back higher up the income distribution and keeping it below a 100% marginal rate to eliminate perverse incentives, you could do that with UBI and tax changes, too, and still have a (smaller) net win.