My primary purpose in quoting the Pucher report was for this conclusion:
You extrapolate from the whole quote a reference to studies which show a strong preference for separated infrastructure and turn it into an argument about preferences between "door zone bike lanes" and protected lanes.Results from aggregate cross-sectional studies indicate that there is a positive correlation between cycling levels and the supply of bike paths and lanes, even after controlling for other explanatory factors such as city size, climate, topography, automobile ownership, income and student population.
I think the Pucher reference to that preference was a pretty simple one. For example, I have two routes to downtown Boston. An 8.1 mile streets route with almost the entire distance with bike lanes and sharrows. The alternative being a 10.2 mile route which includes one mile of back street riding to a separated bike path for 8 miles and another mile of sharrows to my destination. I would answer that my preferred route was the one with the separated path. I do the other one when I'm in a hurry. It doesn't surprise me at all that there is a strong preference for separated infrastructure when placed in that context, which may be different from Professor Dill's study if the choice was "door zone bike lanes" vs "protected lanes".