This is a topic I've tried to get people to think about (usually in the B&W processing and Enlarging forums as it is not strictly an exposure issue). My view, particularly when dealing with long subject luminance ranges, is that relatively mild ZS contractions are helpful, but that more extreme contractions and compensating procedures can result in more loss than gain. I still think that beyond a reasonable range, the goal of "fitting the negative to the paper" is flawed when it comes to the use of applied densitometry methods (ZS, BTZS etc). One must consider the printing process when making a negative exposure/development decision in the field. Developing printing skill and techniques allows one to view the negative slightly differently - as an information source. The goal in exposure and development of the negative then becomes more about recording the most information, not about trying to match the negative density range to a grade of paper.
I'm still working through some thoughts on the Nelson/Simonds paper, Delta-X and print judgement speeds so apologies to Stephen for the delay in posting more. I haven't had enough time yet. Please bear with me.