I wouldn't be surprised if you're right.imo, its not by weight
its % THC vs other cannabinoids.
Of course, if this is the case, the measurement is completely meaningless, since the total cannabinoid content could be less than, say, 0.1% LOL, and the THC %age of that could still be, say, 30%, which gives a ridiculous false impression!
But as I said before, I think that a false impression is exactly what the people that bandy these figures are looking for!
Also (and somehow no-one ever mentions this) the variation in strength between phenos of the same seed batch can be massive (100% or more). So if you measure the strongest ones, you are creating unreasonable expectations as it is highly unlikely that the rest of the plants are at this level. OTOH if you measure an average one, then you don't get the 'desired' figures.
So the whole exercise becomes pretty pointless, if not downright idiotic.