The EM energy per layer and gain computed in derivation was missing the cell weight. (It was OK in rel 21.2.) This has been added back.
To have a bit more information, also the ratio (energy with weight)/(energy without weight) is added as a decoration.