I really really like this statistic, but I have a question about how you came to it.
Is your percentage based on total Darwins?
I have run the same test and came up with different percentages. I took the total profitable Darwins for that time frame and compared it to the total number of Darwins within that timeframe.
For example, as of today, there are 510 profitable Darwins with a history less than 365 days and there are 1,086 with a total history of less than 365 days. Therefore, the percentage profitable is ~47%.
Furthermore, with at least 3 years I found there are 40 profitable Darwins (basically the same as you) and 104 total Darwins showing more than 3 years of data - putting that number at ~39% of them are profitable.
Therefore, it looks to me like you may have compared all of them to the total available Darwins. That doesn't quite seem like an accurate measure to me, but I am sure that I am missing something. I'd love to know more about how you came to your conclusion - because it is an amazing thing to see.