If there won't be any replay protection, miners would risk to waste their power on useless coin. This leads to next predictions: if replay protections will be introduced, we can expect to see some 30-60% switching to B2X at least for short term; if there will be no protection, this number will be much lower, around 10-30%. In both scenarios, Bitcoin shouldn't suffer much, worst case we'll drop 25% and than recover in a few days to previous levels.
You seem very optimistic. How about a 51% attack with a super majority of miners? That's what a hard fork with no replay protection is. Between SPV wallets and exchanges like Coinbase and services like Bitpay, most Bitcoin users would be following the hard fork chain.
After all, most Bitcoin users do not run a full node; they default to a service (like the many signed onto the NYA) or they run a non-validating node. If it comes down to a super majority of miners and no replay protection, we are facing a very messy situation.
51% attack means attacker is mining valid blocks with malicious intent - either a chain reorg to make a double spending or mining empty blocks to stop the network. This means that even full nodes are affected by it, while with hard fork full nodes just reject invalid blocks. SPV clients, if configured properly, also detect chain splits. Online services hold the power to decide for their users what chain to follow, which can be very damaging to either chain, but big exchanges would quickly suffer legal consequences if they'll try to force unwanted chain on their users. So, this kind of attack by hard fork is not as scary as it may sound, and it's definitely not the same as 51% attack.