I doubt anyone is suggesting small blocks forever, but at the moment there doesn't seem to be sufficient evidence of an emergency need to take some kind of drastic measures to increase the blocksize limit on an emergency basis.. and in that regard, the impression and/or creation of an emergency seems to be largely fabricated with a bunch of loud and whiny voices.
I think it was something like a year ago when Hearn predicted that urgent action must be taken or Bitcoin would ...crash. Yes, he said crash. The rationale was something like "ohh nodes will fill their mempools and nodes will start crashing and ohhhh, we need bigger blocks".
For a programmer it's very suspect why he didn't simply write a patch and say "here guys, this will fix nodes from crashing through a parameter that sets mempool size"... Why promote a hard fork for 8-20mb increases (which would be 20-50 times the legit activity of the network) instead of simply patching the software with a mempool limit, as 0.12 does?
The whole crisis / urgency scenario was bullshit, as was the "problem=>reaction=>solution" - in terms of ...proposed "solution".