I've done some actual densitometry tests using "EXG1-2", a formula that should be identical to EXG1, but is made to hopefully be slightly more shelf-stable and a bit easier to use. Mixing instructions:
Part A: -- requires heating
* 80ml triethanolamine 99% low freeze grade (what I have, can be replaced by 85% TEA, 15% water)
* 10g glycin
* Top to 200ml with propylene glycol
* Add 0.1g or "a pinch" of sodium metabisulfite. It will mostly not dissolve. This should help to make use of what water is in solution to be more preserving
* Let settle and cool for 1 hour
* Pour into final container. Leave insoluble powder behind, but a little bit in the final container won't hurt anything. Do not shake container before use
Part B:
* 140ml water
* 60g potassium carbonate, anhydrous
* top to 200ml water
Usage: 40ml + 40ml to make 1L of working solution developer. Everything should be pretty much the same as EXG1 including color transitions.
For densitometer tests, I use 35mm FP4+ exposed using exposure setting number 5 on an X-rite 334 sensitometer. I then use an X-rite 331 densitometer for measurement. Neither piece of equipment has been properly calibrated or rather the calibration it had is expired. Absolute results thus should not be taken with too much confidence and so it's more interesting rather for comparison purposes. The test strip was developed in a hand tank for 10m30s, at 73F. 30s of initial agititation and then 4x agitation in ~8s every minute. Several water washes were used as a stop bath and 3 minutes in TF-4 was used for fixing. Final results were washed under running water for 2 minutes and then put into photo-flo solution. Each density was checked at least twice to account for noise and measurement error. All measurements I'd say are fairly confident within a few percent.
The dmin in D-76 1+1 was measured to be 0.31. The dmin in EXG1 was slightly more purple to the eye and measured 0.33. Using unexposed areas of previously processed films, the 0.33 minimum density seems to be consistent on FP4+ in EXG1 regardless of development time. I'm unsure why the dmin is different, maybe due to the lack of sulfite preventing the clearing of some dyes or something.
The graphs I've made are attached. One was made using absolute density and the other using density after subtracting the dmin of each developer. The results are quite telling. The speed index of EXG1 is actually slightly higher than D-76, indicating that it might be very slightly over developed. Obviously reading the charts though, EXG1 has much less shadow density and thus shadow gradation is much more steep, leading to a more "brilliant" result with shadows further away from highlights. Less highlight compensation seems to be evident in EXG1 when compared to D-76. This steeper shadow gradation may also be an effect that masks some of the grain perception.