Very interesting analysis. However, I had never used this add-in random
number generator anyway - either I use RAND() or VBA rnd().

For the RAND() function, your short test yields a highly satisfactory
result: 10000 unique numbers (XL 2003).
I did the same for VBA rnd(), and had the same result.

My takeaway: It seems to be safe to use RAND() or rnd(), even if I
didn't perform mathematical tests on true randomness (I know they exist
but couldn't apply them).

Thanks to all!
Al