I don't disagree that practicing works, but I'm very open to evidence that homework is an ineffective form of practice.
My wife is a music teacher and frequently objects to the phrase "practice makes perfect" because the kind of practice that you're doing has an enormous impact on the outcomes. Practice works to solidify behaviors, but ill-conceived practice will actually end up solidifying the wrong behaviors.
I think it's likely that it turns out that homework solidifies behaviors that are objectively a net loss for the individuals and society.
I don't think it's a particularly fun thought exercise but I know people enjoy this kind of thing. I feel like I've seen so many decisions being made backed by "research" that have caused significant harm. I think if you come up with a conclusion that significantly diverges from common sense it needs to be replicated multiple times and the experiment has to be fool proof before it really deserves consideration.
* A is in S
* For many B in S, P(B) is true
* Therefore P(A) is true
Step three in this chain doesn't logically follow from step two, so preemptively rejecting evidence against P(A) is irrational.