While I’ve done both (raw from the bottle, and post mixing), IME, it’s only a “rough indicator” at best. More often than not, what you end up vaping post-steep, is significantly different enough that I would never recommend trying to learn flavors by a finger test.
I think it’s extremely difficult to get past the varying amounts of alcohol and/or PG (aka the flavor carriers), never mind the varying sources and “brands” of each.
(in addition, there’s no reason to think that every manufacturer sources the exact same alcohol, much less at the same ratio, etc., and something as simple as the chemical makeup of the carrier could/does color the impression)
I agree that after enough trial and error, that you can use a raw finger test from a bottle to gain a very rough idea on starting percentages though. But, only with low to moderately high potency flavors.
I feel that only SF testing is going to prove useful, but especially so on the stronger flavors (1-2% and less bracket) which I have two brackets for: high strength (1-2%), and ULTRA high (sub 1%).