Don't like the options.  Don't think there should be a hard-fork unless bitcoin is attacked by the china-coin chain.  If they want to fork-off, it is their right.  There's no reason to punish the miners that don't.  While I think a PoW needs to be investigated in the medium to long-term, I don't think it's necessary unless bitcoin is attacked.

This is my take on what would happen if BU forked and took 50% of the hashing power.  It would almost immediately see SegWit being enabled.  That's one thing I would love to see, and I suspect the UASF trigger could make this happen, because you could set the UASF trigger to be the middle of a difficulty epoch.

Let's assume we're talking about mid epoch.  The only thing we can really be sure of, yes?  The average?  For brevity, let's say 1000 blocks.

    1000 blocks : one week. 10 minute blocks.

    1000 blocks : 50% hashing rate. Two weeks. 20 minute blocks.

Three weeks til difficulty reset. Then, because the difficulty calc will include both parts of this (50% full-mine, 50% half-mine), the new lowered difficulty will be 2/3 of the difficulty it was before, with 50% of the previous hashing power. So ~15 minutes blocks.

Off the top of my head, I reckon that would be another three weeks instead of two. Then the difficulty resets to the available miners, and you're back to normal.

So :

Fork happens.

    Two weeks : 20 minute blocks

    ~Three weeks : ~15 minute blocks.

Then back to normal. Assuming miners don't see the potential for huge fees to be had by switching over. Cuz, you know, money. I might be out for a few days or so, but I reckon it'll be about this assuming 50%. Seeing as I think bitmain actually has 50% or so of hashing power, I think I'll be close to right.

It's almost like you want bitmain to fork off. Because blocks will be bigger, so more transactions, but the block creation time will be longer, so less blocks. So less miners creating blocks less often, each with more transactions. Net effect to miners? More total fees in larger blocks and more for each miner. Net effect to users? Same fees for larger blocks that are created less often.

For those five weeks, miners will have a field-day. 50% more profit? Good on em I say.

And if they do attack?  PoW change (to something like Keccak that already has an infrastructure?) and bonza, off we go again, no difficulty reset, and everyone wants to pile into keccak mining.  I really hope we have a real solution to miner centralization instead of kicking the can down the keccak road though.