The results are really impressive, thanks for the detailed explanation. Would be interesting to see what would happen to the transaction bursts, when the latency of the global network starts to kick in. Have you conducted a simulation yet, that would prove it to scale linear?