Ok understood. I thought you meant including or removing fb+fog density could change the shape of the curve (ie slope, curve shape etc).
I have to admit I was not aware it was correct practice to in fact plot gross density. I can see the value in that for a single film. But suppose one is comparing two films on the same graph. Since comparatively (ie from the perspective of the printing paper), Dmin must be printed through, why wouldn't we want to remove Dmin from the data? For example I know that Ilford's Delta 100 typically develops to Dmin of ~0.3 and Kodak's TMX develops to Dmin of ~0.2 (using my densitometer). Why wouldn't I "normalize" that out of the analysis?