Swedish people in general are also more likely to be much taller than Chinese people if you prefer to use height as the example, or have a higher percentage of blonde than etc.... The point being there are going to be regional variations based on genetics/breeding.

No the overall BC would not have shifted. Certainly it has in specific regions, making it possible for the TR to breed more Channelers and stronger Channelers and potentially the region where the AS picked up these 1000 to breed fewer and/or weaker. But in the overall scheme the Curve is the Curve, RJ was speaking of the present day when he was quoted on the 62.5% and he made no caveats about current day strength etc ...

You can believe that 1000 is not a tiny sample, but no statistician would ever tell you that a regionally confined sample representing such a small % of a population is reliably projectable as a sample. Things we can assume about the WoT populacation. There are a dozen or so cities of significant size representing in the ballpark of 8-10MM people. With the population being heavily skewed outside of the cities we can assume a total population in the Westlands of maybe 50M. This means in just this population there should be roughly 250k women with the ability. Factor in Seanchan Aiel, Sea Folk and Shara, and we may be looking at a channeling population of over 1M and from vastly varied regions and genetic circumstances. Sharans from what we know actively breed the trait so it's possible they have increased the % of the population to 3% or maybe higher if that is possible. Seanchan effectively eliminate all Sparkers from the breeding pool, but nearly all female non-Spakers are found and are given places of high standing and likely breed normally. We've seen more strong women among the Aiel and SF than maybe we should have (I'll leave that to speculation since the author tended to put stronger Channelers at the forefront of his story lines).

Net result is we don't have anywhere close to a representative sample with what amounts to about 1% of 1% and that number is heavily regionally skewed. It would be like trying to project something on a global scale based only on the population on Topeka Kansas. Some things will likely work out, but the regional bias will have a heavy effect.