My best luck was with a phenidone-hydroquinone-sulfite mixture. Phenidone is great in the shadows, but hard as heck to get much contrast out of. It has an advantage over most othe agents in that it is not much affected by bromide, so you don't really need a replenisher. I used it until it got to looking so bad my conscience hurt, but never wore it out. I used 0.65 grams phenidone, a bunch of hydroquinone, mixed in a 10% sulfite solution. The bunch was 8 or more grams. It seems that, like the Metol-hydroquinone combination, the activity of the developer does not change perceptibly after the ratio of hydroquinone to phenidone exceeds about 10:1. The hydroquinone, in the presence of sulfite, replenishes the phenidone much better than sulfite alone, or at least that was my interpretation of some data I found in a book in the NASA library some time in the 60's. It seemed to work. If you want to give it a Q&D trial, 1/4 tsp phenidone, 4 or 5 tsp hydroquinone and 4 tbs sodium sulfite in a liter of water will develop film in about the same time as D-23. The phenidone will dissolve in water easier if you mix it with a little alcohol or glycerine first. Dry, it's about like talcum powder. It just floats there. Maybe even a few drops of liquid dish washing detergent would work as well.
The next step would be phenidone-ascorbate-sulfite, or even phenidone-ascorbate-borax. See the side article on experiments with metol-ascorbic acid-borax. It includes phenidone as well.