David A. Goldfarb said:
Just to follow up on Sandy's comment, I think the desire for "bulletproof" negs with more shadow density as well as increased contrast probably has to do with some of the films that people who like contact processes often use. If you use a film with a long toe and high Dmax (Tri-X, Classic/Forte 400, and I'd guess Super-XX might work this way as well), overexposure pushes the whole range up the curve, off the toe, and into the straight line portion, resulting in better shadow separation, and as long as the film doesn't shoulder at the highlight end, this can produce better results.
I am sure that is why some people overexpose film but it does not always achieve the desired end, in fact more often than not it does not.
Overexposure to push shadow density into the straight line part of the curve works really well with some film/process combinations, Super-XX and AZO coming to mind as a superb example.
Unfortunately it does not work nearly so well with other combinations and is actually counterproductive with many. For example, TRI-X has a very long toe but an upward flaring curve in the highlights so if you overexpose to put important shadow density in the straight line part of the curve it makes highlights very difficult to print in AZO (or other silver gelatin papers) without getting blown out. This perhaps explains why TRI-X has never been very popular with AZO. On the other hand the flaring curve in the highlights is great for Pt./Pd. because it compensates for the highlight shouldering of the process so it makes sense to overexpose TRI-X with Pt./Pd.
And of course you don't want to overexpose films that have very short toes (HP5+, FP4+) or films that have almost no toe (Delta 100, TMAX-100, TMAX-400) with any papers, irrespective of the shape of the paper curves, because with these films the shape of the curve is such that overexposure simply adds to printing density without achieving any enhancement in print quality.
In short I don't recommend overexposure for most film/process combinations unless you really understand how the film and paper curves interact with each other. But folks who already know that don't need someone else's recommendation anyway, now do they?