It's a lot closer to UBI than existing systems. The Finnish case is an experiment where a random sample of unemployed folks can be paid for some time even after taking a job. It explores the impact of the very incentives you're talking anout.

(And yes, the lower-middle class will probably still be net UBI recipients if only barely, because this helps with smoothing out incentives at the lower end. You don't want to have a sudden benefit cliff, or else a breakeven point that's infeasibly high.)