Note that I said "I could go on and on", these are just a few examples of what could affect a test ride. There are many more variables I can list out. You cannot possibly control for everything.
EVEN given that you have controlled for absolutely EVERY external variables, how do you know that you simply won't err? In other words, barring ALL other variations, how can you be sure that what you felt to be say "stiffer" or "more comfortable", is actually true, from just one ride?
There are no guarantees in life, but your odds of getting things right are increased when you control what you can rather than just taking it on faith.
You can tell a lot from one ride, and one thing you know for certain is that if you liked your experience, there is at least one configuration that will make you happy -- frankly it doesn't matter why. Besides, even a n00b can tell if they prefer race or comfort road geometry and notice marked differences in the rides you'll get from different frame materials.
If the test ride doesn't matter for a n00b, geometry doesn't matter. Normally, I'd conclude that if such is the case, more people would buy recumbents. Except 'bents typically cost more. That's undoubtedly the main reason you don't see more of them...