Diafine came out when film emulsions were still fairly thick and could absorb a fair amount of the first solution. I haven't used it for a long time, but I think present methods of stand or minimal agitation development are at least as likely to get the desired results. That's just an opinion, but the last time I compared D-23 with the D-23-borax split developer, the characteristic curves were indistinguishable. The extra step was not worth the effort.
It might prove something if one were to compare Diafine 2-bath with a mixture of Diafine A and B. If that one, which will not be I, got both characteristic curves to have the same contrast index, one might find everything else the same as well.