Just remember my post wasn't a shot at you, just suggestions/comments on the procedure. Your first post contains no scientific information, only that you used the Xipp dyno...nothing about a lab and controlled environment with calibrated equipment, etc.
Perhaps you are right, but I haven't found anything close to your findings with...100 or so springs. Not to say it isn't true, but you still left the notion that you aren't measuring the rate at the same compression. I mean, my 40lb springs could be 16.87lbs at 5% compression. However, all four of them are within 1lb of each other, and I can live with 2.5% variance. I can because using a matcher I can get them as close to identical as practical.
There is no doubt that each spring is different, I'm just skeptical about 15-20% per spring.
Perhaps you should buy your springs somewhere else, or change brands.
Food for thought...and I don't think it is trivial or anal to be this detailed. 5/100/lap on a track that gets 32-33 laps/race is ~1.6s, and that's a big lead.