Well, the other 15% of miners now control the network if every merchant/user/service/exchange/etc did not agree with the hard fork the miners had, and the difficulty will adjust downward on that side of the fork to compensate for the real network mining losing a substantial amount of hash power. The original 85% of miners that voted on and adopted this miner-voted hard fork would basically be mining nothing since merchants and users aren't accepting their chain.
I did not know that one merchant or user or service or exchange could forestall a hard fork approved of by
all remaining parties. Quite the concentration of power.
Universal veto.
miners provide security for economic incentive, and it is in their best interest to make a profit on "work" (hashrate). They have TOTAL control over hardforking the blockchain by the very definition that only a miner can create blocks. Merchants, users, exchanges, services that do not mine blocks ave zero ability to create blocks and must follow the rules set by miners in order to have a transaction confirmed or marked valid.
obviously, merchants/users may not agree with a fork implemented by miners and may stay on the unforked ruleset, with transations mined by those miners. If they decide to send the same coins on both sides of the fork, that would also be fairly easy to do. But if a majority of hasrate in on the new chain, the older chain becomes vulnerable to hashrate attacks and using it for transacting becomes unsafe or unpredictable.
In this situation, fees become important. If all the transactions (and thus fees) are on the minority blockchain, miners would have economic incentive to return to that ruleset.
Again, it's just like if I decided to start mining blocks with a 1,000,000 BTC block reward. Sure, I could mine as many as I want, even if I have 99% of the network hash rate.... but no one is going to accept them, so it's pointless to do so. A hard fork doesn't need universal consensus, but it needs a super majority of miners, merchants, and users for sure... otherwise it doesn't make sense.
As a simple example (and I'm in no way affiliated with these entities, just well known examples): Let's say miners decide to hard fork, but Bitstamp, Coinbase, and Bitpay decide not to follow along and continue using normal clients. Well, no one mining coins generated by the hard forked chain could sell coins on those exchanges, but the miners who stayed on the normal chain could easily do so (and would be making a bunch more since their side difficulty would have dropped). It doesn't matter transaction volume, fees, whatever... if the miners can't sell or use their coins... what good are they?
So again, miners have zero control over the implementation and adoption of a hard fork. They sure have control over execution once the community agrees on one, but miners have no authority to make such changes on their own without the rest of the community (users/merchants/etc).
It may seem difficult to grasp and accept, but it's the way it is. And, IMO, is quite an excellent design. If it were any other way miners could decide to hard fork and do completely crazy things. A hard fork could be virtually any change to the system, even ones that are completely irrational. I mean, if it were that simple then miners could have decided a while back something crazy like, every transaction included in a block entitles the miner to 1 BTC in additional subsidy. Or, any output not spent after X days is spendable by anyone. Etc etc. This is why a HARD fork is not possible with JUST miner involvement. Essentially, any hard fork converts bitcoin into a new coin, requiring new clients and all for EVERYONE in order for it to be utilized... just like if you decided to mine some random altcoin, you can't treat your altcoins as bitcoin and vice versa.
---
On pool news, I've been making steady progress on the new changes that are coming soon to the pool. I'm running over my initial time estimates a bit due to other unrelated work running a bit into overtime, but I'm doing my best to catch up and get things finished here.
I'll try to do a better update soon, once I have some more to share.