> which purports to gauge a study's rigorousness as a simple percentage

That's not what it does.

Some of the studies showed a causal relationship clearly. Some showed merely an association with a plausible argument about why it was mostly causal. You can decide what proportion of the unproven effects to include.

In reality, there's more causal effects than those that are proven, but not all of these measured associations are causal.