I worry this entire poll is methodologically flawed, as many people have pointed out...
People who want a change are active participants. They're going to work to get the change they want. They're going to be paying attention to things regarding that change. Voter turnout will be very high among that empassioned constituency.
People who don't want a change may not have their voices heard, because they may not be paying attention to the issue. Its already been decided as far as they know, and they may not realize its even up for debate again until the vote is already over. In which case you've solved nothing, you've just switched the positions of both sides, and then the forums will be filled with an empassioned instant-transfer constituency pushing vocally to have the change reverted. At the very least, the vote should have been announced in advance so people could be aware of it and expecting it...
We shouldn't have to actively affirm that we want something left alone. It should be assumed that most people who don't express an opinion are satisfied with the existing status quo. Unless voting is mandatory for everyone, you're going to have a *strong, strong* sampling bias in these results.
edit: ESPECIALLY with the stated benchmark that the "instant" vote has to be an overwhelming majority to cancel the change. That's not fair. It should be the other way around - it should be the "delay" vote that needs a strong majority to demonstrate that a plurality of players want a delay, to compensate for the poll being skewed towards that side. An equivocal result should indicate a conservative outcome, ie keep it the way its been planned thus far.