Yes, that could well be the case. Given that the results are instantly apparent, in practice one would presumably just sit down with the light and a box of capacitors and, within a very few minutes will have ascertained, by trial and error, what was needed to produce the desired effect...As various sources in the past have indicated, the results is by no means linear. Typically I have had to adjust to much lower inputs than the brightness would have indicated. My control method was pwm so in the present case the results may be different, but initially I'd lean towards the smaller values myself.
Kind Regards, John