Kodak doesn't say that. In any case, it isn't that D-76 was/is unsurpassed for emulsion speed/shadow detail...
WRT Bill's post above, yes, metol is a good start, but not the entire story.
Actually, I tend to live in the past. 'Unsurpassed' was regularly used by Kodak back in the 60s and 70s (and maybe later as well). NOW they simply state 'excellent' shadow detail. Perhaps this is a bow to, and affirmation of, how extolled is Xtol developer in our current darkroom world.
I also did careful tests with Dektol diluted about 20x and that shadow detail was just as good as with D-76. But, again, Polydol was deficient by about half a stop (although its highlight separation was better than with either D-76 or highly diluted Dektol, or D-72). Configured as a portrait developer, this trait for Polydol just might not have been happenstance, as the somewhat superior highlight rendition with Polydol would have been a distinct, though subtle, benefit for such employment.
I think that D-23 would not do any better with shadow rendition as one poster, Bill Burk, seemed to infer: this metol-only developer seems to perform normally, albeit gaining gamma (contrast) does take more time.
I wanted to posit this question because there have to be tangible reasons for such differences. Your comments are revealing, although, perhaps, not yet definitive.
But I do take issue with the nomenclature comment posed by Matt King (correct me if I am wrong). Film speed is a determinant of different things for different folks. The history of such determination is diverse and even controversial. For me, however, it is that emergent shadow detail,
in conformity within a consistent, predetermined, overall contrast index, which definitively rules the speed of a given film. - David Lyga