> from ontological knowledge (naming and categorising)

The example clearly is discussing naming and not categorizing, which is precisely why Feynman’s anecdote is not fully applicable to the Linnaeus system.

> They have unity

I agree. That doesn’t mean the naming system can’t be improved, it just means unity is a goal and might be harder to achieve than before. This is what I meant about goals and agreement. If the new system were better enough, and unity was sought in advance, achieving unity might not be too difficult. It has happened before…