I will take the time to listen to the samples a bit later when things are somewhat less noisey around the house. However, if I may, I'd point out a few flaws in your quiz. The wine tests are somewhat well known and used in many debates which center on subjective perceptions vs objective "facts". In other words, the perceived desire of the well heeled to prove they are superior to the rest of us. My career has been mostly involved with high end audio and that has taken me into numerous ... er, "discussions" which involve whether what has been claimed can actually be perceived and repeated. Or, is it all "snake oil" and objective "facts" are all the facts that count and the amplifier with the lower distortion and higher measured wattage will always be the superior product? Due to the dollar amounts asked for certain pieces of equipment and the highly personal situation in which music is perceived by the individual listener, there are grand arguments waged between the various camps who at times have their very existence - if not their reputation and their following of loyal supporters - on the line.
The ability to detect small and at times momentary dissimilarities between components or even something as seemingly inconsequential as the directionality of signal flow in a speaker cable is what "high end" audio is based upon. The basic question is, why would anyone spend the money for any product they did not perceive to be "better" than another? The answer to that question is to resort to double blind tests which are meant to determine whether perceptions are honest and consistent or simply a function of expectation bias and the commonly known placebo effect ... and its near relative, the "no-cebo effect". If you believe, you are 90% of the way to perception or lack there of.
High quality blind testing is a very difficult thing to accomplish. Small errors which would not present themselves in another situation are heightened in their effects once a subject is asked to determine which is which and which is superior. An almost ancient article written by the man who created subjective listening audio reviews back in the 1960's is included here;
One point which stands out in any such test is the approach taken by the test subject once they have been put on notice their responses are being graded. In such a situation, the listener mentally switches gears and is no longer listening to music for pleasure or even as a critic of the performance of the player. Another portion of the listener's mind is switched on and now the listener is being asked to determine which is which amongst what have been established as unequal values. In other words, you've already established in our mind there are differences between the samples and we now focus on the differences and not on the music. We are literally testing ourself as to whether we can determine which sample is which and no longer are we listening as if we were not under such a test. This variation in skills is not how we normally listen to music. What we come to is a test of the test subject rather than a test of the equipment. It is no longer about whether subjective differences exist and it becomes very much about whether any one listener has the ability to perceive those differences we have been told exist. This is likely the single most difficult value to overcome in designing a non-biased test and where most test designers fail to think ahead.
Secondly, yes, by its very design, there is significant information lost in any of the numerous MP3 formats. If someone were to come into my shop and want a demonstration of a high quality sound system, they wouldn't be given that demonstration using MP3 material. If they told me they were intending to use MP3's as their primary music source, I would never suggest they invest heavily in quality audio equipment since the subtle improvements found in superior audio equipment will be lost in the source player and format. This doesn't even get into the variables presented by offering such samples on a computer sound card system. Which playback format will be used? Will the samples be audition through headphones or speakers? What is the quality of the system being used? And, will the lower quality of an average computer system played back through less than transparent equipment favor one sample over another?
This is not a discussion of the possible losses found in the transfer from one digital format to another but simply a statement of how MP3 was created to supplant quality with quantity. The MP3 format was created through the use of audio masking technology which throws away information when it would (possibly) be obscured by louder sounds in various frequency bands. Attack and decay are inferior to both WAV format and the real thing since there are less bits carrying the same amount of information from the original music source. In short, MP3 is a compromised format and should not be the choice of decisions based on somewhat critical impressions of quality.
Please don't take this as a criticism of your quiz but rather just another view of the testing procedure you have chosen. It is extremely difficult to design a test of any one's subjective perceptions and not run afoul of the many pitfalls which are inherent in such tests. In the end, while some results might be interesting to discuss, if your objective is truly impartial and meant only for your own enjoyment, then that is where your test must end. No real world conclusions can be made from such tests.
I still own my laminated Yamaha classical from the late 1960's. I also own what I would consider several rather nice solid wood instruments. When playing the Yamaha I am still taken by its overall tone which is warm and mostly unobjectionable. Its sins of omission are largely a matter f it does not possess sins of commission. It is, though, a very difficult guitar to play when it comes to projection, attack, sustain and (not so) subtle variation in emotion. It is a far more difficult guitar to play effectively than are my solid wood instruments. If a test subject's priorities are geared mainly towards tone and tone alone or the player is not quite as skilled at bringing forth those varied expressions of musical performance and the listener is not aware of their existence or the music example selected does not lend itself to those elements of performance, there could easily be a conclusion reached which would not favor the higher quality instrument. I would liken this to those clients who desire a high end audio system yet are completely unfamiliar with the sounds and perceptions involved in a live music experience. If your reference is your car stereo, what you will find in a high quality home system will not be very satisfying. This is even more the point with a hand built instrument which can involve virtually every variable imaginable from the quality of materials and construction to the humidity on the day of the recording.