The chlorine loss rate would be a percentage so 0.5 out of 4.5 is somewhat over 10% though your resolution is only 0.5 ppm when using a 10 ml sample so this isn't really a fair test. With the higher levels of FC you were seeing 17 vs. 22 drops, a difference of more than 20%. Still possible, however. I just find it surprising that the swirling of the Speedstir is that vigorous. It's certainly very helpful at getting a better reading in the CH test where it requires much more thorough mixing and where doing it manually can take a much longer time having to swirl between each drop.