I did read the article, and the effect of highlight separation by pyro although well documented is still a matter of perception. Nobody has proven conlusively that pyro does "better" than any other developer when used appropriately.

If anything again pyro is more effective with pt/pd as it blocks more UV light as explained in the article you mentioned. Arentz and Herbst have done this study where pyro is compared to a "normal" developer and it was concluded that a normal developer is just as capable of producing the same results. As I stated you have to fit the negative to the curve and the pyro stain does not make the paper "better", azo is simply better able to accomodate the separation by its own nature, not because of the pyro negative. IOW if you have a highlight that is 1.87 and another that is 1.9 azo would be able to separate them as shown in the article, but if you have a 1.9 and a 2.0 highlight no matter what you use, pyro, d76 or anything else, the 2.0 would still show as paper white. The fact that the pyro stain can better diferentiate between closer densities does not improve the paper. The paper is only showing what it can do, no more no less, no matter what developer has been used. I would say that the tonal scale of azo is actually relying more on the paper developer used than on the negative. I think Michael has proven this, azo in any other paper developer is just another silver paper, amidol is what lends it the better reproduction qualities.

From my experience in printing in silver I can tell you that given a negative that contains the same reproduction scale of the paper, all the tones will show in the paper. Reproduction is dependant on the paper, not the negative.

This is the problem of doing step tablet tests for something as subjective as photographs. Which reminds me if you go back and look at the azo article you will see that the print reflective density range is exactly the same for both the pyro and the d 76 negatives, this cannot be unless you have some give somewhere, the d 76 neg appears to have better separation in the dark tones. I am sure given a little bit of more work the d 76 negative could have shown better separation in the highlights also.