This might sound like nitpicking but I believe it presents an important conceptual distinction. Flare doesn't change the film curve. How can it? The illuminance range from the subject striking the film has the lower tones compressed by flare producing the equivalent effect of extending the toe. You can see how that works in Quadrant I, Camera Image / Flare Curve in the examples I attached (which are from a program I wrote).
Originally Posted by markbarendt