i guess 95% would be ready within 3 months.
Depend on circumstances, if we get huge capacity problems and people not able to confirm transactions (remember 50 seat bus cannot serve more people no mater how much they pay, you need another bus or send them to competetion), then 28 days is plenty of time as well. But if SegWit become usefull and helps enought with the capacity problems, 3-9 months seems reasonable for 2MB hard fork activation time.
But planning ahead instead would be better, it would be only advantegous for Bitcoin if we had such 2MB buffer already, no need to streess the market (thus price) and sending the people to altcoin competetion.
But no, altcoin lovers and centralized hub operators are two biggest fractions in Bitcoin Core and they need to get rich with altcoin prices going up or limiting Bitcoin onchain capacity for their still not existed centralized hub operations - really disgusting for Bitcoin holders and true supporteers.
Why bother when we know that holders have no power over bitcoin network at all? We all know our views are not important.
Only miners, hash whales and bitcoin companies could have something to say here.
I agree, Satoshi concensus = hash decides. This is the only provable fair way to decide. Obviously miners must keep in self interest price and where to sell or spend, so both holders and companies influence miners decission, developers too because they might be important for future code improvements. Funny some Bitcoin Core developers still thinks they are the most important and others have no power to change Bitcoin rules, saying Bitcoin Core = Bitcoin
