For me the AI got to the point where it just became rather dogmatic, and wasn’t able to see the shortcomings in its’ argument that I was pointing out to it, and then declared itself the winner. So given that I’d say it was able to provide a rather high fidelity simulation of the real world debating experience.