I preffer if the next month conference gets consensus about the one time blocksize increase, and whenever the blocks become filled again in future another blocksize limit adjustment has to be made (hopefully not taking 1 year of pain in future again 
)
)What would you say if the consensus at the conference was a design for scaling up that did not involve altering the blocksize? That's apparently what that conference is about (all solutions, not just one engineering approach).
Ideally the cap wouldn't even be needed. It's there because there is a vulnerability in the form of forcing tonnes of work on everybody in asymmetric fashion and with unpredictable dynamics. When this is cheap and easy, the system just needs a safeguard or it's trivial to attack inexpensively.
If miners could agree to sustainable criteria for tx inclusion following some standard and feather-forking the miners who don't follow it, the cap would not be necessary.