Okay, text is terrible for my communication style, I apologise. 20% of HASHPOWER is feasible and distinct from the sybil risk that node count might include. If a BIP110 compliant block is found immediately after activation, that is the longest chain of work. Legacy chain nodes will verify because it is consensus valid. If a BIP110 block is not found immediately after activation and a non-compliant is found that block maybe rejected by (as of now) 10% of the network. In that scenario, if 2 Compliant blocks are found, that legacy block will be orphaned. That legacy miner is now incentivized to avoid that scenario. Those are a non game theoretical look at a how it could play out. A game theoretic prisoner's dilemma scenario would be: After BIP110 activates, a single large miner with 20% or higher hashrate defects from legacy and simultaneously subtracts 20% of the total hashrate from the legacy chain and finds a block with 100% validation by of all nodes. The rest of the mining actors would now compete with a 20% higher difficulty while also facing block rejection from BIP 110 nodes. which incentivizes them to prepare for that eventuality and be ready to switch over to BIP110. If not, they risk mining on a dead chain.