Part of economics SHOULD be a science. Let me clarify...
Economics comes from a long and noble tradition dating back 2000 years, and being deeply researched since around the so-called Scottish Enlightenment. Political economy, as we now call it, wanted to see the disciple holistically...embracing the political and social (even historical?) forces that undoubtedly drive the variables in our models.
Marshall's revolution rightly tried to flung economics into the Comptian tradition of socio-dynamics...looking for an identifiable force to drive all observable phenonomen. Its those seeds that lead us today to scientificize (to coin a word) even love, dating, abortion and many other aspects of seemingly uneconomic life. Yes, we drew on the same motivations and heros that the physicists drew on (and many of us like me are failed physicists).
So yes, we -- and most vehemently among the AEA - strongly adhere to the methods and appearance of science. The use of objective data. Impartiality. The teasing away of complexity through modelling. We hide our uncertainty and misgivings in pages of obscure details (which might be a scientist's attention to detail in so much as the hope to bury the sources of spurious regressions). We still use ancient looking tables and regression panels, when more attractive, illustrative methods could (and should) be used.
We train our students to be hyper-numerate and quantitative (like scientists)...but always spin a good tale, and offer real-world recommendations, like the Physiocrats (failed to) do.
So, yes and no. Many of our peers - especially in the AEA (no disrepect to my brothers and sister) cling to the accoutrements of science (try getting great European or Asian economics in the AEA's flagship publications). Yet, many of us, reject thus entire paradigm...calling it "the pretense of knowledge" (ie Hayek). They...and even I...wonder why we are "discovering" the biases and cognitive shortcuts (and awarding Nobel Prizes for these 'discoveries') which the psychologists and fish monglers of yesteryear well knew. We are good atlabelling everyday stuff with fancy monikers (from aversity, voracity and veracity effects to signalling).
Yet, as the biologists and paleontologists teach us, sometimes just labelling something makes it useful science.
PS - sorry, I just saw the link. I thought you wanted something more discursive. A perfect illustration of the problems of our discipline in deciding what counts as a survey and useful data :)