come on gmaxwell.. even you know better then that..
i think many have quoted you "Gmaxwell: Segwit is not about saving space for plain full nodes"
so even you know that segwit and other features are going to cause 'big blocks'

come on gmaxwell.. even you know that segwits 1.8x capacity (roadmap guessed 1.5-2x yet further investigation expects 1.8x) means 1.8mb of real fullnode data.
come on gmaxwell.. even you know that things like confidential payment codes turn a 500byte into a over 750byte average size.
turning 1.8mb blocks =2.7mb (without any capacity increase, just bloat instead)
then add on the extra bytes not in the tx but needed to link weakblocks in the block adds more, along with other flags too.
so while staying with the 1mb hard limit there is actually 2.85mb of real data per block once all features are combined from the roadmap

then when the final hard fork happens to turn the 1mb hard limit.. to 2mb.. the butterfly effect is
2mb traditional = 3.6mb segwit
3.6mb segwit = 5.4mb segwit+CPC
5.4mb segwit+CPC = 5.7mb segwit+CPC+all other features and flags

if all you can do is throw an insult and not actually give some maths, code, or statistics .. then you are not doing anything to convince people to your way of thinking.
so how about try to give some actual numbers that can be traced.. that explains the real extra bloat that will occur .. oh and to close a loophole. make sure you numbers are for the actual real FULL data of all the roadmap features used together and transmitted between FULL NODES.. try not to meander the numbers to hide them by talking only about pruned no witness lightclients.. be honest and give numbers for FULL NODES

that is my challenge to you..
no insults. but instead. actual byte for byte maths, stats and details of the bloat of using all of the roadmap features.