Unfortunately "most" users want larger block sizes to decrease wait times while "most" miners want block sizes to stay at 1 MB so difficulty is doesn't increase, all their ASICs still work, and so they get more money (IIRC people can pay for fast-lane processing).
If the devs pushed the 4 MB change the miners wouldn't move and because of this no transactions on the 4 MB bitcoin would actually work. If the devs don't push 4 MB then all of the users leave to another coin and bitcoin drops in value.
Talk about rock and a hard place.
Almost all of their resources would be the same - the vast majority of their electricity bill is spent solving the block hash compute problem, the resources to actually process the transactions and send them are small compared to that.
The basic logic holds, that you cant make money if nobody uses Bitcoin.
The blocksize has been increased in the past, without major upheaval. There is some risk, but staying with outdated software forever also has its own risk.
We should make the single hardfork count - eg. an agreed schedule to increase blocksize over time to keep pace with moores law was proposed as a way to avoid future recurring hardforks.
I think the issue is to manage the risk well - telegraph and explain the roadmap so there is certainty for all parties. Gain consensus, but understand perfect consensus is impossible.
How big should we make the block so that only big banks with specialized hardware can afford to run the bitcoin network.