The claim in P1 is a general assertion: legal systems don’t do as good a job as hoped. The claim in P2 provides some support for that general assertion by observing one reason why earlier systems have fallen short of expectations. The claim in P3 gives further support by observing why more recent systems are still less useful than hoped.

The claims in P2 and P3 aren’t full arguments; they’re just claims. (The claim in P2 could be considered a sub-conclusion supported by the rest of P2, while the claim in P3 is simply an assertion of fact.) And both claims support the general assertion in P1—neither refutes it.

This gets the direction of support backward. “Entails” means “implies.” But the assertion in P1 doesn’t imply that the assertions in P2 or P3 must be true. The fact that legal reasoning systems have fallen short of expectations doesn’t tell us that such systems must specifically have problems with interpretation or with matching cases to precedents. Rather, those two specific problems are independent observations that the author makes, and she uses those observations as support to explain why, exactly, legal reasoning systems have issues overall.

The observations in P2 and P3 don’t refute the claim in P1. They support that claim.

The claims in P2 and P3 aren’t incompatible. They’re observations about two different types of legal reasoning systems, and they both support the claim in P1.