My consensus network PoW design places no theoretical limit on the TPS. Physics will still place a limit else where (i.e. aggregate bandwidth of the entire internet), but that limit won't be at the layer of the consensus network. In short, I don't see any practical limit yet, but I caution there has been no peer review. And I've been pretty much overloaded while making this discovery, so it is possible I've overlooked a flaw. I don't think so. The one flaw I'm contemplating is how to set the minimum transaction fees to avoid a Sybil attack while retaining decentralization of that adjustment. But Bitcoin has the analogous dilemma in that if block size is unlimited then attacks are possible (unless also set a minimum transaction fee) and these constants are not set with a decentralized process. So at least my design seems to have the potential of a great leap forward, even if the Holy Grail of decentralization will never quite be entirely devoid of politics.