I'm pretty much with you on this, though I'm in favor or changing the 1MB limit. I don't like the proposed solution of 20MB + exponential increase and I would not fork to it (so count me as 'anti' for the poll). The proposal creates problems.
It is a sort-of simple proposal, (which is in its favor), but I think we can do better.
I'm actually in favor if raising the limit as well under certain conditions:
- There is a well defined end-goal and a well studied (and obviously, open) path to getting there
- There is a need based on the above.
It just makes ZERO sense to do a Hail Mary and get a tiny fraction of the way to anything meaningful just because it 'feels' like the right direction. Bitcoin is already important enough to where this type of thing does not make any sense, and under the quasi-management of the Bitcoin Foundation there is an appalling lack of transparency which totally sucks. The little bits that do sneak out, like the discussions about 'red-listing' are not very heartening...though they do have a high potential to explain the current direction since increasing the exclusivity of operators is a necessary first step.
We aren't in a crisis, and there may not be a crisis for quite a while. Even having the horrible circumstance of queueing transactions (this already happens for low/no fee transaction), is not the end of the world for Bitcoin, but it will limit growth and adoption.
So since we have time to improve the proposal, this is what we ought to be doing.
If we are changing from a static to a dynamic limit, it ought to be sensitive to the need for increase rather than exponentially increasing. This could be either too high of a limit (and creating the potential for vastly increasing network costs unsupported by Bitcoin economy), or it could be too low (and we end up back where we are now).
I would support either a sensitive increase, or a static increase (which just kicks the can down the road but caps the risk), but not a dynamic one that has exponentially increasing risk.
I'm 10-times more negative about dynamic adjustments than I am about simplistic block size for scaling. Throw in the towel at that point. They are extraordinarily difficult to get right in the best of circumstances, and that is far from the case in Bitcoinland where everyone and their brother are trying to game the system for fun and profit (as the design correctly anticipates.)
Bitcoin is utterly lacking in any viable feedback mechanisms to make dynamic adjustments tenable. Just to set fees have struggled mightily with this. Adding such mechanisms into the mix would necessitate a fairly significant overhaul (my gut sense of things.) If Bitcoin had some sort of 'benevolent dictator for life' then _maybe_ dynamic feedback would work _sorta_ for a while. But if Bitcoin had such a thing it would not be Bitcoin. The window of opportunity for dynamic feedback probably ended at about the time Satoshi announced the whitepaper. That bird has flown. The best bet is to address the need by careful design of autonomous semi-auxiliary systems in my opinion.