Phil Carmody wrote (at inordinate length):

> ln(1+1/d)/ln(b)

and I checked 25 thousand LucFib cofactor primes,
greater than a million, in base b=10:

d # fraction Benford=log(1+1/d)/log(10)

[1, 7869, 0.313] 0.301
[2, 4456, 0.177] 0.176
[3, 3107, 0.123] 0.125
[4, 2342, 0.093] 0.097
[5, 1977, 0.078] 0.079
[6, 1665, 0.066] 0.067
[7, 1360, 0.054] 0.058
[8, 1255, 0.049] 0.051
[9, 1107, 0.044] 0.046

Looks good to this heathen physicist!

David