Have you seen this?
http://it.slashdot.org/story/12/02/14/2322213/998-security-for-real-world-public-keysThankfully to us Satoshi did not choose RSA for the private/public key algorithm of bitcoin!
This is important nevertheless. Ok, 2 per thousand is statistically very low, but the fact that
all these vulnerable keys can be gathered by any skilled enough attacker is quite troubling.
I wonder how fast would the bitcoin development team be able to work out an algorithm migration if a similarly dangerous vulnerability were to be found on ECDSA or SHA-256 (these are the algorithms used for public/private key and hashing in bitcoin, respectively, right?)