The Bitcoin Core developers were never part of the agreement to hard-fork and increase the block size - they've said it was a terrible idea all along at least on the time scale proposed. I don't think they even knew about it until it was publicly announced. Also, SegWit increases the block size to 2-ishMB by itself (it's complicated, because older nodes need to think blocks are still under 1MB, which is done by stripping out parts of SegWit transactions when sending to them). This means that in practice the hard fork gives a block size of 4-8MB.
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.