Here's a quote from the SciAm article: "Technically, that equation was t/log(t), but for the numbers involved log(t) is typically negligibly small."
Huh?
Maybe I'm missing context, but that sounds like O(n) or Ω(n).
t/log(t) is 'closer to' t than it is to sqrt(t) as t heads toward infinity.
e.g:
4/log2(4) = 4/2 = 2
sqrt(4) = 2
2^32/log2(2^32) = 2^32/32 = 2^27
sqrt(2^32) = 2^16
I think this means that while Log grows to infinity, it does that so slowly that it can often be treated as if it were a coefficient. Coefficients are ignored in big O notation, hence the negligibly small comment.