It seems the conversation on the 'potato' taint died over 2 years ago. I'd love to revive the conversation to see if there are any new observations or resources that one or many would like to share. That said, here is what I have:
In the cup/brew batch, it seems like the instances of the potato taint are diminished with the institution of 'quaker yanking'. I'm using quaker liberally here, as I am applying it to the ever most slightly to severely under-roasted coffee. Given that Peter G said, "Flotation of cherry before pulping helps, as does densimetric sorting of finished coffee," and Tim C. said "I suspect that the defect or the resultant lightness of beans or the internal issues of the bean," it shouldn't surprise that the under-roasteds are the culprits.
I am wondering if other roasters have found that the problem rests within the quakers, or if there are other physical features that we could be looking for?
Re: identification of under-roasted coffee... We have a set up at the gimme roastery that utilizes both fluid bed and drum roast technology. I have noticed that I have an easier time identifying these under-roasted and quaker, thereby potato tainted coffees from our fluid bed roaster than from our Probat, due in no small part to the kind of surface color uniformity you get from windchill. I am not arguing that this type of uniformity, or the means, are a blessing, but, however, it seems makes the task of identifying lower density coffee via blond color easy as pie.
If the answer to the above question is yes, other roasters have found that the problem rests within the quakers, to what extent and ease does one 'quaker yank'? keep it clean, folks.