Satoshi set the 1MB cap as a temporary measure, but he predicted that users would get "tyrannical" against raising it because of not being able to run nodes and so on.
Also, Satoshi did NOT have the current understandings of what different blocksizes would mean, he did not predict mining pools, he did not predict a lot of things.
ALL the experts point to a hardfork in 3 months being stupid, we should evaluate segwit for a longer time then take a more informed decision.
I totally agree here. My post was not about advocating for a 2MB hardfork in November. I also would like a longer evaluation period before any hardfork.
But there is - in the Bitcoin community - a difference between the 21 million rule and the blocksize. In the "blocksize war" there are notable different opinions, with a relatively large group of Bitcoiners agreeing to each position: The "extreme big blockers" that don't want any limit and would even be OK with gigabyte-large blocks, "moderate big blockers" (2-8MB) and "small blockers" (at most, 1 MB with segwit). In contrast, the 21 million supply cap is nearly unanimously agreed on - so it's much more difficult to change it with a hard fork. That's what the post was about.
One thing that is not emphasized strongly enough on this topic is that it is very difficult to envision a scenario in which there would be a positive reason to increase the coin supply beyond the current plan.
A reason that could reach a bit of support would be the deflationary nature of Bitcoin, once we reach a supply between 20 and 21 million and the block rewards are worth less than the coins lost in addresses whose private keys aren't known anymore. There are disagreeing views about deflation, according to the economic schools, but there could be a group with some support on that reasoning. Only that I don't think it will ever reach consensus (what would be necessary to raise the cap in an
undisputed hard fork).