Thanks to Nick W, dee arnall, gandhi, Theo and Anonymous Poster for pointing out how the Washington Post arrived at the 1 in 8, and obviously, to Mike Rivero for reposting my brief and obviously superficial analysis.

I had the same thought as I was falling asleep last night. Guess I should have waited to write my analysis. It has been too long since I studied polling, but I think DesMondu is correct that each question should be considered independently.

I dispute Jeremius's contention that anything over 55% is "solid majority," despite the broad use of that terminology in regards to elections. A shift a few percentage points, which could happen in just a month's time, could bring this majority into minority, especially with a margin of error of 3-4%.

I still stand by my #1 claim - that the WashPost is underestimating support for withdrawal - but I'll have to add the caveat that it's somewhere between 12.5% and 41%.