Here dichromate is not working in its photosensivity capacity but as an oxidizing agent. At lower exposures, it acts to oxidize the Fe(2) back to Fe(3), cutting the toe part of the curve, as Koraks said. As far as I know the azo moiety is not oxidizing so it would not be expected to work in the same manner as the dichromate. There are other oxidizing agents and I did some inconclusive experiments with a couple of them - potassium permanganate and hydrogen peroxide with the classic cyanotype. There are also sodium/potassium perchlorates. But I have not seen them used successfully in the literature, so makes me wonder why. Nitrate ion is also oxidizing and said to act like dichromate in the Simple cyanotype by Mike Ware. In that case, the nitrate is formed in situ from reaction of ferric nitrate, citric acid, and ammonia.
You could perhaps try another way to clean up the highlights - selenium or sulfide toning. For example, I toned POP with KRST which increased the Dmax as well as bleached the highlights. KRST didn't work well for salt which bleached both ends, but Kala Namak did. The effect is unique for each paper/chemistry/process combo, so you will have to experiment and see if such an approach works in your case.
:Niranjan.