To restate the issues so those who missed the finer details of the problem:
1) This problem of D76 leaders being weak (looking more like coffee stains than being opaque) and Xtol dying within a couple of weeks has occurred repeatedly over the last 8-9 years with the fresh lots coming from Freestyle (Arista 76), 5+ gallon bags; BHPhotoVideo (Kodak D76), approx 15 gallon bags; and 10 5L Xtol bags from Freestyle and BH. All of these were mixed with the Kroger DW exclusively.
2) No liquid developer has ever exhibited less than an opaque leader when developed to manufacturer's recommendations, including, but not limited to, 3 bottles Tmax, DDX, 3 bottles of HC110, 3 bottles of Rodinal 500ml. Kroger DW was used with all these developers over the past same time period.
3) Wet prints made with D76 developed negs are always FLAT compared to the liquid developers, pretty much consistent with the weak densities in the leaders and highlight densities. Recent measurements using a spot meter as a densitometer confirm 3 stops difference between a D76 leader and a Rodinal leader, both developed to mfrs. specs. Using one poster's suggestions, I was able to determine that the D76 leader developed normally only reached D=2.1 where the Rodinal leader reached 3.1+ when developed normally. Negs exposed using Sunny 16 at box speed or 1/3 stop less and developed to Kokak's specified times, straight and 1:1. Minimum 250ml or more per 80 sq. in. Negative material includes both 100 and 400 speed film.
4) Bottled drinking water used for mixing D76 for the last 8 rolls have show a leader opacity differential from Rodinal about 0.5 stop. I just got back from vacation in S. Utah and have over a dozen rolls to process and will see if the rest of the gallon and the new batches work as advertised.
5) I don't own a densitometer but have employed my Pentax spot meter as a makeshift densitometer.
6) Bottom line question: Why should D76 exhibit 3 stops less density in its leader (and Xtol die anywhere from 3 days to two weeks) when mixed with Kroger DW when no liquid developer exhibits less than virtual opacity? Why would bottled drinking water with pH of 7.8 when mixed with D76 have caused the leader density to increase by 2.5 stops?
I don't know the answer; I only know the bottled drinking mixed with D76 has improved results comparable to results in the 90s when I lived elsewhere and used tap water and worked with fully developed negs.
My curious nature now wonders why?