Well it's probably better than just instituting a program and hoping it works. Ideally, all government programs should be subject to measurement and re-evaluation.
The choice does not have to be between "do a bad study" or "do no study at all".
A proper study would be permanent and self-contained. Participants would need to give a portion of outside income to the program to simulate higher taxes and/or inflation.
Results can start to be worthwhile early on. But the life-altering decisions people make over a two year period when guaranteed a subsidy that runs out at the end of those two years are not particularly likely to be the same as the life altering decisions people make over a two year period when guaranteed that subsidy for life.
(If you wanted to really understand all the dynamics, you'd probably want to wait more than a generation to understand how subsidy recipients' decisions to raise children are affected and whether those children make systematically different life choices.)
A proper study would be permanent and self-contained. Participants would need to give a portion of outside income to the program to simulate higher taxes and/or inflation.