I empirically tested the mathematical theory of payment channel networks.
I simulated 1 million payments on the 2-party channel LN, compared to a 3-party channel Network (same total cap)
Supported amounts increase by almost two orders of magnitude! Payment Failure rates halved!