I have heard that some people pick through and screen their cocoa beans one by one to remove all the uneven ones, flat beans or ones that don’t look just right. Do I have to do this? You don’t mention it at all.
Do you have to do this? In a word, no, you don’t have to do anything you don’t want to do. Further, I don’t do it except in rather rare cases where I have very questionable beans (with a note that I do my best to never offer those said questionable beans). But that does not mean I don’t do something to clean up certain beans before they fully turn into chocolate. For me, it’s more of a zen approach. A balance of effort vs results. Here is what I mean by that.
In the first case, you can take the time to rather laboriously spread out your beans and pick through what you think are the questionable beans. When I’ve done this, I have indeed found that they were easier to process, I had less flats and husk getting into my winnowed nibs (which I still picked out) and …. compared to the chocolate where I did not pick through the beans one by one, I could not say there was any perceptible difference in the resulting two chocolates. That right there has me (call it lazy or frugal or smart or efficient ) not wanting to put in extra effort for no benefit that I can discern.
In the other case, using the zen approach, I let the process do the work for me. If I see obvious debris, I pick it out as I am going along. During the roasting process, in my perforated drum, any broken nibs and small debris natural drop out. Screening without an extra step. The agitation will also break up any already cracked (hence questionable) beans and drop them out. Next the cracking and winnowing process further pulls out a lot that isn’t quite right. It’s designed to crack and separate beans that are of similar nature (you are picking and screen for similar beans after all). I’ve found if you have over fermented beans, they tend to shatter more and get carried over (via the Aether or Sylph) with the husk. Under fermented beans don’t release their shell as well or fracture into larger pieces and often have their husk still on the nib. If I notice a lot of these larger pieces I will indeed at this stage pass the nibs through 1/4” hardware cloth (through trial and error, this size discriminates just right I’ve found). Clean nibs of the appropriate size (and presumably preparation) drop through, and the larger pieces are left on top. So, sure, I screen, but it’s a fast process, not a picking through one by one. If there is a lot left, I will run this quickly through the cracker and winnower again (instead of running the whole batch through again). The result is ‘cleaned up’ nibs without any extra steps or time. And it separates on a more objective set of parameters than your visual observation.
So that is my take. Let the process work for you. Don’t coddle the process.