I tried to redo the experiment from the article and developed Tri-X400, TMY-old and TMY-new in Kodak D76 1+1 to the same CI in a Durst Filmetta. There was no difference between TMY old and new besides a 10% difference in dev. time needed. Tri-X had a bit coarser and sharper grain. This exactly matches what Kodak says about the garnularity and difference between the old an new films.
I don't now what they did in that article but as long as nobody is able to succesfully redo it (the gold-standard of any scientific result) I'd say they simply mixed up the samples they got back from the lab where they had them developed.
If anyone is interested I could upload my samples in the tech-gallery.
Stefan
Sounds like a reasonable explanation, Stefan; i.e., "I'd say they simply mixed up the samples they got back from the lab where they had them developed."
For Kodak's current take on TMAX granularity, See Kodak's F-4026 tech pub on the Kodak website.
TMAX 100 Diffuse rms Granularity is given by Kodak as 8 (extremely fine).
TMAX 400 Diffuse rms Granularity is given by Kodak as 10 (extremely fine).
For TRi-X granularity, See Kodak's F-9 tech pub.
TRi-X Pan TX: Diffuse rms Granularity is given by Kodak as 17 (fine).
TRi-X Pan Professional TXP, TXT: Diffuse rms Granularity is given by Kodak as 16 (fine).
For comparison see Kodak tech pub P-255:
For Kodak Tech Pan developed in HC-110 (Dilution D) the Diffuse rms Granularity is given by Kodak as 8 (extremely fine).
For Kodak Tech Pan developed in Technidol Liquid the Diffuse rms Granularity is given by Kodak as 5 (Micro fine).