@Jason Stoddard...here's an idea for when the blind testing begins again.
I'm a physician and cancer researcher - I think it would be fascinating for us to collect some real data on blind listening results. If you collect the data, I'd be happy to do the statistical analysis and write up the results (you can be first author). I suspect we could get the results published in a peer reviewed journal, perhaps one of the otology or sensory rags. I'm not aware of anyone who has published this kind of work, and I think it would be fascinating. Data collection could also happen at one of the headphone meets, perhaps a CanJam...
First question we'd have to answer is how large the sample size should be, and how much of a difference we'd want to detect. I'd defer to you on this one...if only 5% of people can detect a difference between two pieces of gear, we'd need a much larger sample size than if 50% can determine the difference. You would know better than I where that number sits.
Second, what you would want to test? Headphones? DACs? Amps? All three? Again your call. Probably best to test something more easy to detect like tubes vs. solid state amps as opposed to cables or even DACs. Headphones would be a good subject as well - though not as relevant to you at Schiit.
It's probably cleanest to just test whether each individual can detect a difference between two pieces of gear, as opposed to whether they can identify a specific piece of gear, though you could probably collect both sets of data and look at everything.
Who would participate? "Golden Ears" only? General population? Schitt gear owners only? Again your call.
How rigorous would you want to be? If we're shooting for publication, we'd need to keep the data very clean, which means controlling for all variables, cables, source material, bitrate, etc.
We'll then do a very basic statistical analysis, get a p-value, and have an answer whether the differences detected are due to chance or whether they are real.
PM me if you're interested - happy to help out in any way needed.