This is a good description of the both the problem (legal reasoning systems don’t work well) and the explanation the author gives for that problem (these systems have issues with interpretation and matching cases to precedents).

Too strong and too general. The author doesn’t argue that computer programs in general have little value. In fact, she notes several long-standing uses of computers for legal professionals at the start of P1. The problem she raises is just with legal reasoning systems in particular. And she doesn’t quite suggest that such systems have “little value”—she just says that they’re not as useful as was initially hoped.

Anti-supported. The author is critical of case-based systems because they fail to solve the underlying problem of how to get computers to reason through how precedent applies to different cases.

Anti-supported. The author notes in P2 that doctrine-based systems won’t work “in the foreseeable future” and in P3 that case-based systems have an “intractable” (i.e., difficult to solve) problem.

Descriptively inaccurate. The author argues that computers have difficulty matching cases to legal principles or rules (see P2), and she never suggests that computers offer insight into reasoning flaws.