The "compromise" framing of "we gave you SegWit, now you have to give us a hard fork" is pretty terrible too, because the two aren't symmetrical - SegWit makes hard forking less urgent, but hard forking to increase the block size makes the SegWit improvements more important because they're designed to make larger blocks work better. Even Bitcoin Cash copied one of the key SegWit improvements, the transaction hashing change, in very slightly modified form for this reason.
More recently a large group of players in the space came together and agreed to what has come to be called the New York Agreement (NYA) This said that they would support Segwit and a 2MB hard fork for bigger blocks. Core and Blockstream who largely controls Core have ignored the NYA and worked against it. They are pretending like segwit has support w/o the 2MB block increase when it doesn't and they are going around doing some truly cringe worthy stuff to tray and scuttle the 2MB part of it.
[0] https://www.reddit.com/r/btc/comments/6uw5kf/why_does_core_r...
No, a couple of developers did. I think every core developer not in the room went "wtf, no".
As a result, it's very likely to see another hard fork to 2 MB in November, so there will be 3 chains with a shared transaction history, and most like all will claim to be the real Bitcoin -- personally I don't think debating which one is the "real" one is worthwhile, just use the one you want, but eventually it's likely that one chain will get the majority of hashing power and have a much higher price per token than all the others.