What you are describing as the so called asymmetric advantage of BIP 110 is really just a property of any soft fork that tightens consensus rules while remaining within the valid set for old nodes. Old nodes can accept blocks from upgraded miners as long as those blocks do not violate their current rule set. That means that in the short term the upgraded side can potentially pull in hashpower from miners who are not explicitly enforcing the new rules simply because the chain tip they see is valid in both rule sets.
However this is not a magic bullet. The moment the upgraded side has less total hashpower than the non upgraded side it becomes a probabilistic race. The idea that 45 percent could reliably win a long term fork is flawed unless that minority is consistently luckier or has some coordination advantage. Over time the side with majority hashpower will extend its chain more quickly and the minority chain will struggle to stay in contention. The asymmetric part only manifests in temporary events after a split when the upgraded chain wins a block race and can pull non upgraded miners onto its side for the next block or two. Once a block with invalid transactions from the perspective of upgraded nodes appears the split resumes and the advantage disappears.
The real lesson here is that activation thresholds and miner coordination matter more than any supposed asymmetric property. A soft fork with marginal hashpower is fragile especially when contentious. The safest path for BIP 110 or any consensus change is clear majority support before activation to avoid these chain tip oscillations and the uncertainty they create for both miners and users.
What you are describing as the so called asymmetric advantage of BIP 110 is really just a property of any soft fork that tightens consensus rules while remaining within the valid set for old nodes. Old nodes can accept blocks from upgraded miners as long as those blocks do not violate their current rule set. That means that in the short term the upgraded side can potentially pull in hashpower from miners who are not explicitly enforcing the new rules simply because the chain tip they see is valid in both rule sets.
However this is not a magic bullet. The moment the upgraded side has less total hashpower than the non upgraded side it becomes a probabilistic race. The idea that 45 percent could reliably win a long term fork is flawed unless that minority is consistently luckier or has some coordination advantage. Over time the side with majority hashpower will extend its chain more quickly and the minority chain will struggle to stay in contention. The asymmetric part only manifests in temporary events after a split when the upgraded chain wins a block race and can pull non upgraded miners onto its side for the next block or two. Once a block with invalid transactions from the perspective of upgraded nodes appears the split resumes and the advantage disappears.
The real lesson here is that activation thresholds and miner coordination matter more than any supposed asymmetric property. A soft fork with marginal hashpower is fragile especially when contentious. The safest path for BIP 110 or any consensus change is clear majority support before activation to avoid these chain tip oscillations and the uncertainty they create for both miners and users.