Now I guess this where the big blockers don't agree. I personally do believe 1 MB blocksize wouldn't be enough forever, sooner or later it needs to be increased. Big blockers believe a soft fork implementation to increase blocksize wouldn't be sufficient in the future and I guess Segwit + 2x increases the base block limit to 8 MB.
But hard forking the network three months from now would be suicidal. I am not dragging myself into the politics of bitcoin, just my opinion. Core has been working on bitcoin from the beginning. I don't know if Core would agree upon a hard fork or not. If yes then the Core developers and Jeff Garzik team of developers could work together to create a robust code through efficient planning and multiple testings and then implement it one year from now.
In my honest opinion I do believe bitcoin should scale with an equal blend of off chain and on chain implementations. Off chain solutions might be a bit harsh on miners, once all the bitcoins have been mined, incentives would be transaction fees only, a necessity for a PoW protocol, and off chain solutions would keep them in check.
Nope.
Playing around with miners fees and what they may mine in the future is stupid. If miners don't get the fees, they won't mine and Bitcoin will fail. You are also making the same assumptions that the miners in 20 years time will be the same miners now. Punishing today miners at the expense of future miner is folly.