Why I think the "consensus-level transaction fee" proposed by @jessquit is a bad idea: https://read.cash/@Jessquit/a-simple-elegant-way-to-solve-instant-transactions-on-bch-350be853#comment-98e01510
Elaborate schemes to fix double spend to the n’th degree is a distraction. I don’t see it as a pressing real-world problem. Maybe one day, but not now.
Let's say this 51%+ hashpower is invalidating any block including a CashFusion transaction (active censorship). CashFusion transaction fees rise, as they seem not to get confirmed.
- Block subsidy don't really protect the network against censorship (although it helps against double spending). - Proof-of-stake is less censorship resistant than
| Honest miners deploy new hashrate in order to mine these transactions (economic incentive). At some point, censoring hashrate falls below 50% and honest miners can build the longest
With the deep reorg protection on BCH, this is more complicated but it can still be done. You also still need to have a decent fee level, so a lot of users should be ready