If you think your opponent is arguing for miracles, it is likely that you are misinterpreting their claims.
Too many liberals are far too confident that the reforms they advocate would have been better than PRWORA and would be better than continuing current policy, because they tend to not seriously consider unintended consequences. Reforms that make receipt of welfare relatively more attractive will tend to draw more families onto the rolls and thereby run the risk of preventing them from benefiting from the advantages of employment. Or it will discourage planned and responsible childbearing by reducing its costs.
Relatedly, many critics of reform are too dismissive of paternalistic anti-poverty policy. “Don’t people know what’s best for themselves?” ask liberals and libertarians who assume that if work pays, people will naturally work. There is much to be said for this view, undoubtedly, but even if most people know what’s best for them most of the time, some do not at least some of the time.
Furthermore, policy shouldn’t necessarily strive to give people what they think is best for themselves in a world where wants are potentially unlimited and someone else is paying. Some people value leisure more than work at the margin, and in that case we have to ask whether working taxpayers are obliged to support the leisure of those who could work but do what’s best for themselves by not working. The same may be said of people whose childbearing decisions are what’s best for them; none of us is entitled to do what’s best for ourselves and expect others to bear the costs. These particular ways of evaluating the success of welfare reforms are, in a sense, moral rather than economic. But conservatives hold these normative views, in part, out of the belief that work and responsible childbearing may benefit children economically even if they do not make parents happier in the short run. For that matter, they may make society better off economically in the long run by promoting economic growth.
In the end, the conservative resistance to weakening the tough provisions in the 1996 law stems from the belief that — regardless of whether there might have been a better way — welfare reform improved the lives of the poor when compared with the old system. We get very nervous about departing from a model that a lot of evidence suggests was better than the status quo.