I think what GMW is saying that I did a test which I ran for months in secret playing with the publisher name and it had not statistically significant impact on anything. My sample size of course was limited to my own books/market/etc.
Use your own name, and forget it.
