That reminds me that Gmax's main point earlier this year, when pressed, was that small block pressure was the only thing that could incentivize efficiency improvements to be implemented, and that if needed the raise could be put through in a few days/weeks.
If it really would be that quick and easy to raise the limit, both the need to raise it preemptively and the need to ensure it is plenty high largely fall away. Gmax as optimizer of code wants there to be more incentive for people to implement what he is good at, and of course for BS's projects, and he seems to want more recognition of how much the optimizers contribute.
Though those lines of argument may be disingenous, it would waste time to mistake the facile arguments of the forum gadfly true believers for those of the Core devs and the Blockstreamers, even though it is remarkably disappointing how little some of their arguments rise above that level, as they are not strictly the same.