This is simply not possible on bitcoin unless you have wallets holding money from many people. At which point you no longer own your wallet. There is no way Tx

This is simply not possible on bitcoin unless you have wallets holding money from many people. At which point you no longer own your wallet. There is no way Tx of average humans would significantly cancel out within the time of a block. Thats why LN just makes the time to settlement "infinite" and that's why it partially increases centralization again to cancel out more Tx. If that was the goal it could have been implemented in layer 1. Anyway all that aside, lets just assume LN would work as perfectly as the devs hope it will do one day. Its sill way worse than existing FBA solutions that can handle most of this on layer 1 directly and already have fully working second layer (payment channels) implemented. Even if bitcoin and LN could eventually provide a similar good solution, whats the point if it already exists and why would anyone assume the existing solutions stop evolving while bitcoin and LN slowly catches up? If we assume LN has reached all its goals in 10 years it also means the goals are outdated by 10 years. We can arguing about what LN may be capable of doing one day but its never gonna be a valid argument against solution that deliver today.