That’s a great two cents!
Just remember, though, that the people who do very official research do appreciate this forum as a very powerful “incubator” of sorts. The amount of anecdotal feedback and mini-tests here do serve some very useful functions. It often tells the people who design more rigorous studies WHAT to study further.
We’ve also gotten results over the years beyond what we originally thought we would.
As just one example, we’ve found users here who unusually adept at this type of testing. They must have bias correction strategies internalized in their own heads. (Ie - their minds are always adjusting to compensate for any preconceived notions they might have) Some can even figure out what pheromones are in a mix, to an eerie degree – even down to their proportions – far before the mix’s contents are objectively revealed. It’s mindblowing for me to see that sort of tack-sharp, perceptive feedback. Even I can’t do it nearly as well as the few I’m thinking of, so it blows me away every time I see it.
While these people are few and far between, we never would’ve found them without a massive forum like this. In a way some of these people are more impressive than some of the mixes, so that’s a positive “side effect” in a way. 
Many didn’t even start this way. The forum actually encouraged and perhaps enabled them in some way to develop these unusual social awareness and testing skills.
Maybe it’d be good to think of this forum like an apple test garden designed to break free from weak monocultures.
If you’ve ever seen such gardens, you know they must be massive since each planted apple seed generates a completely unpredictable result. Braeburn apples don’t just spawn Braeburn apple seeds, for instance. Hidden in the seeds of every single apple, is every single apple’s genetic material. If you planted six different seeds of from the same Braeburn apple, you would get six different apple trees. Few produce precisely the exact result everyone hopes for, but a type of evolution still occurs. Learning definitely happens on many levels. When a good end result does develop, it tends to be an unusually hardy one that can weather a number of trying environments.
Such a garden has many experiments going on at once and many “gardeners” designing them and both gardeners and testers testing the “fruits” of their labor.
Later on, the more rigid research then acts as a set of very necessary “sanity checks” against the gardeners and tester’s biases (their hopes, fears and outright delusions.)
The sanity checks we’ve been working with in the background are less ambiguous than even the most rigorous types of social testing. Most social studies still have far too many variables even when you strip the variables bare. For really “hard” data, we look directly at how the brain itself responds to a stimulus. While we’ve been doing this with relatively basic technology since the beginning to evaluate the molecules we offer here for testing, we’re moving more in the direction of ultra modern testing involving both fMRI and SST. Expensive stuff, but ultimately the results will be worth it.