Spurious Correlation - "A false presumption that two variables are correlated when in reality they are not. Spurious correlation is often a result of a third factor that is not apparent at the time of examination. Spurious comes from the Latin word spurious, which means illegitimate or false."
Nothing to do with the math...
If you get enough legitimate data points with sufficiently high correlation, you can say with increasing certainty that the correlation is real, which is the math. If the correlation is significantly over .5 or so, spurious correlation due to a third factor such as a common divisor can eventually be statistically ruled out. Many times (but not always), you can work through the variables being analyzed to determine whether they are ratios, have a common divisor, etc., so that third/common factors can be statistically removed or better variables can be chosen.
The bigger risk in my mind is that people for the most part know what gas they are diving, and can be predisposed to giving feedback consistent with their expectations. The cost and scope of a significant double-blind study with proper variables would be interesting to me, at least.