I have added a new, more detailed code in
http://www.elitetrader.com/vb/showthread.php?s=&threadid=81457
The code searches for the optimal Zig change n, from n=3% up to n=10% [you may enlarge this range but for n>10 you will have very few Peaks].
For the last 10 years of the ^NDX, the 31% of the retracement ratios d is 0.50<d<1.00 for n=10%.
If you try to narrow the range, only the 19% is between
0.25<d<0.50 for n=9%.
More detailed range gives worse results, only the 12% of the retracements is
0.60<d<0.70
for n=10%.
Some of them may be considered as 0.618, but is is just the 12% of the total retracements.
This is poor and it is impossible to use it in any serious decision making.